design and application of an instrumented pendulum device for
TRANSCRIPT
University of IowaIowa Research Online
Theses and Dissertations
Summer 2012
Design and application of an instrumentedpendulum device for measuring energy absorptionduring fracture insult in large animal joints in vivoBryce DiestelmeierUniversity of Iowa
Copyright 2012 Bryce Diestelmeier
This thesis is available at Iowa Research Online: http://ir.uiowa.edu/etd/3285
Follow this and additional works at: http://ir.uiowa.edu/etd
Part of the Biomedical Engineering and Bioengineering Commons
Recommended CitationDiestelmeier, Bryce. "Design and application of an instrumented pendulum device for measuring energy absorption during fractureinsult in large animal joints in vivo." MS (Master of Science) thesis, University of Iowa, 2012.http://ir.uiowa.edu/etd/3285.
1
DESIGN AND APPLICATION OF AN INSTRUMENTED PENDULUM DEVICE
FOR MEASURING ENERGY ABSORPTION DURING FRACTURE INSULT IN
LARGE ANIMAL JOINTS IN VIVO
by
Bryce Diestelmeier
A thesis submitted in partial fulfillment of the requirements for the Master of
Science degree in Biomedical Engineering in the Graduate College of
The University of Iowa
July 2012
Thesis Supervisor: Professor Thomas D. Brown
Graduate College The University of Iowa
Iowa City, Iowa
CERTIFICATE OF APPROVAL
_______________________
MASTER'S THESIS
_______________
This is to certify that the Master's thesis of
Bryce Diestelmeier
has been approved by the Examining Committee for the thesis requirement for the Master of Science degree in Biomedical Engineering at the July 2012 graduation.
Thesis Committee: ___________________________________ Thomas D. Brown, Thesis Supervisor
___________________________________ Yuki Tochigi
___________________________________ Nicole M. Grosland
ii
2
ACKNOWLEDGMENTS
I would first like to thank Drs. Yuki Tochigi and Thomas Brown for their devoted
support and guidance throughout my valuable time spent in the Orthopaedic
Biomechanics Laboratory. I truly appreciate the experience acquired under their
supervision. Thanks to Tom Baer for creating the solid foundation for my work to build
on. Thanks to Dr. James Rudert for his manufacturing contributions, passing on a portion
of his machining expertise, and his knowledgably thorough discussions. Thanks to Dr.
Jessica Goetz for her willingness to help me prevail through the challenges with the
motion capture system. Thanks to Julie Mock and Tammy Smith for their seemingly
effortless abilities to complete any necessary administrative task. Thanks to Dr.
Anneliese Heiner for her material testing assistance and laboratory safety efforts. Thanks
to Dr. Donald Anderson for his presentation advice and window watching
encouragement. Thanks to Dr. Doug Pedersen for helping with computer issues and
making blunt, yet honest, presentation comments. I would also like to thank my fellow
graduate students for all their help with this work, and their ability to convert any day
into an enjoyable one. A special thanks goes to Justin Boltz for his assistance in
refreshing my memory on Pro/ENGINEER. I am also grateful for the additional support
that I received from my friends and family. Financial support was provided by NIH
CORT Grant P50 AR055533, DOD Grant W81XWH-10-1-0864, and DOD Grant
W81XWH-11-1-0583.
iii
3
TABLE OF CONTENTS
LIST OF TABLES ...............................................................................................................v
LIST OF FIGURES ........................................................................................................... vi
LIST OF EQUATIONS ..................................................................................................... xi
CHAPTER 1: INTRODUCTION ........................................................................................1 1.1 Objective .....................................................................................................4 1.2 Blunt Impaction Models .............................................................................5
1.2.1 Borreli et al. Pendulum Device ........................................................5 1.2.2 Tochigi et al. Drop Tower ................................................................7
1.3 Intraarticular Fracture Models ....................................................................8 1.3.1 Furman et al. Intraarticular Fracture Model .....................................8 1.3.2 Backus et al. Drop Tower Device .....................................................9
1.4 Limitations of Previous Models ................................................................11 1.5 Energy Absorption Measurements ...........................................................12
1.5.1 Charpy and Izod Impact Tests ........................................................12 1.5.2 Abdel-Wahab et al. Izod Impact Test .............................................13 1.5.3 Van Zeebroeck et al. Pendulum Device .........................................13
1.6 Moving Forward .......................................................................................15
CHAPTER 2: PRIOR DEVELOPMENT OF THE INTRAARTICULAR FRACTURE MODEL ....................................................................................16 2.1 Offset Impaction Technique .....................................................................16 2.2 Tripod Anchorage System ........................................................................18 2.3 Original Pendulum Device .......................................................................21 2.4 First Application of the Original Pendulum Device .................................22
CHAPTER 3: DEVELOPMENT OF THE PENDULUM DEVICE AND DATA COLLECTION TECHNIQUE .......................................................................24 3.1 Initial Pendulum Instrumentation, Modifications, and Additions ............24
3.1.1 Overview ........................................................................................24 3.1.2 Rotary Potentiometer ......................................................................25 3.1.3 Sled System ....................................................................................27 3.1.4 Data Collection Components ..........................................................30
3.2 Final Pendulum Instrumentation, Modifications, and Additions ..............31 3.2.1 Rotary Potentiometer ......................................................................31 3.2.2 Linear Potentiometer ......................................................................32 3.2.3 Low-Friction Sled and Linear Bearings .........................................35 3.2.4 Adjustable Mounting Mechanism ..................................................37 3.2.5 Stopping Mechanism ......................................................................38 3.2.6 Sled Lever .......................................................................................40 3.2.7 Coil Spring ......................................................................................41 3.2.8 Data Collection Hardware ..............................................................42 3.2.9 Pendulum Arm Modifications ........................................................46 3.2.10 Stability Modifications .................................................................46
3.3 Data Collection Technique .......................................................................51
iv
4
CHAPTER 4: VALIDATION STUDIES ..........................................................................53 4.1 Foam Specimen Validation Study ............................................................53
4.1.1 Experimental Methods ....................................................................53 4.1.2 Experimental Results ......................................................................55
4.2 Motion Capture Validation Study .............................................................57 4.2.1 Experimental Methods ....................................................................57 4.2.2 Experimental Results ......................................................................60
CHAPTER 5: MATERIALS AND METHODS FOR IN VIVO FRACTURE SPECIMENS ..................................................................................................65 5.1 Preparation for the Live Animal Impaction ..............................................65 5.2 Pre-Impact Surgical Procedure .................................................................66 5.3 Intraarticular Fracture Creation ................................................................67 5.4 Post-Impact Surgical Procedure and Animal Care ...................................69
CHAPTER 6: RESULTS & DISCUSSION ......................................................................71 6.1 Live Animals #1-4 ....................................................................................71
6.1.1 Intraarticular Fracture Creation Results & Discussion ...................71 6.1.2 Energy Measurement Results & Discussion ..................................73 6.1.3 Further Modifications .....................................................................76
6.2 Live Animals #5-11 ..................................................................................76 6.2.1 Intraarticular Fracture Creation Results & Discussion ...................76 6.2.2 Energy Measurement Results & Discussion ..................................77
6.3 Limitations and Potential Solutions ..........................................................79
CHAPTER 7: CONCLUSION ..........................................................................................80 7.1 Intraarticular Fracture Creation ................................................................80 7.2 Energy Absorption Measurement .............................................................81 7.3 Future (Funded) Research .........................................................................82 7.4 Future of the Energy Absorption Technique ............................................82
APPENDIX ........................................................................................................................84
REFERENCES ..................................................................................................................86
v
5
LIST OF TABLES
Table
1. The characteristics for the progression of coil springs, including the respective spring constants, maximum compression displacements, and total energy capacities. ......................................................................................................41
2. Energy measurement results from the potentiometer and motion capture data for each foam specimen, where KE is the kinetic energy, SE is sled energy, and EA is energy absorption. ....................................................................................59
3. Foam specimen penetration depth and specimen deformation measurements. ........60
4. Energy measurement results for Animals #1-4. ........................................................72
5. Energy measurement results for Animals #5-11. ......................................................76
A1. Dates and types of impaction tests that were conducted with the pendulum device. .......................................................................................................................83
vi
6
LIST OF FIGURES
Figure
1. The cost of treatment of arthritis in the United States per year is $114.5 billion, and the distribution between rheumatoid arthritis (RA), non-posttraumatic OA, and posttraumatic OA is demonstrated. Image adapted and reproduced from Brown et al. .....................................................................................2
2. Percentages of fair and poor outcomes of acetabular and tibial plateau intraarticular fracture treatments. Images adapted and reproduced from McKinley et al. ...........................................................................................................3
3. Close-up view of the components of the Borrelli et al. pendulum device. Pendulum arm (A), foam (B), trolley assembly (C), load cell (D), impactor (E), polyethylene block (F), and X-Y table (G). Image adapted and reproduced from Borrelli et al. ...................................................................................5
4. Tochigi et al. drop tower device in the surgical environment. The clamping release mechanism (A), the drop arm (B), and drop mass with accelerometer inside (C) are labeled. .................................................................................................7
5. Cradle and indenter for creating an IAF in the mouse knee. Indenter and custom lower limb cradle for creating tibial articular fractures (left). Schematic showing alignment of indenter with tibia during loading (right). Image adapted and reproduced from Furman et al. ....................................................8
6. Preload and impact alignment of closed porcine knee model. Posterior view of knee potted in aluminum fixtures and held with six springs (left). Dashed arrow represents indenter alignment to impact a knee without fracture; solid arrow represents the alignment to impact a knee resulting in an intraarticular fracture (right). Image adapted and reproduced from Backus et al. .........................10
7. Schematic illustrating the difference between the Charpy and Izod impact tests. Image adapted and reproduced from Callister .................................................12
8. Schematic representation of the pendulum arm with the optical encoder, force sensor, accelerometer, and impacted body (“fruit”) are indicated. Image adapted and reproduced from Zeebroeck et al. .........................................................14
9. Drop tower fracture device (A), close-up of a mounted ankle specimen (B), three human ankle fracture patterns (C), schematic of porcine distal tibial fracture technique (D), and a single porcine fracture pattern (E). The fracture patterns between the human and porcine cases are comparable. Note the severity of fracture in the porcine fracture. Images adapted and reproduced from Tochigi et al. ....................................................................................................16
vii
7
10. Schematic of the offset impaction insult (A), the introduced stress-rising saw cut technique (B), three fracture patterns from moderate energy impacts (40 joules) (C), and three fracture patterns from high energy impacts (60 joules) (D). The fracture patterns between the compression and offset impactions are comparable. By comparison with Figure 9, note that the severity of fracture in the offset impactions is less than in the compression impactions. ............................17
11. Schematic (left) and radiograph (right) of the tripod device-to-bone anchorage system. Images adapted and reproduced from Tochigi et al. ....................................18
12. Three fractures created using the tripod anchorage system in the porcine tibia (left). Note the general consistency of fracture line location. Representative time-tracking confocal microscope images at a near-fracture site produced using the tripod anchorage system in the porcine tibia (middle) and using the compressive impaction technique in the human ankle (right). Cells labeled green were alive, while red cells were dead. Arrows indicate the edge of the fragment. Image on right adapted and reproduced from Tochigi et al. ....................19
13. Chondrocyte viability results (2 days post-impact) in impaction-fractured joints vs. non-impact osteotomy controls. Dispersion bars indicate 75th- and 25th-percentile values. ..............................................................................................20
14. Original Orthopaedic Biomechanics Laboratory pendulum device. The pendulum arm (A), release mechanism (B), support beams (C), specimen table (D), and pendulum drop mass (E) are shown. ..................................................21
15. Porcine distal tibia fractures created using the pendulum impaction system. ..........22
16. Picture of the general components of the instrumented pendulum device. The angular position (Angle) of the pendulum arm measured by the rotary potentiometer and the linear displacement (x) measured by the linear potentiometer are shown. ..........................................................................................24
17. Side view (left) and front view (right) of the rotary potentiometer attached to the hinge of the pendulum arm. ................................................................................31
18. Original linear potentiometer that was attached to the low-friction sled. .................32
19. New linear potentiometer that was attached to the low-friction sled. ......................33
20. Schematic of the original low-friction sled (A), linear bearings (B), and shafts with support rails (C). ...............................................................................................34
21. Schematic of the new low-friction sled (A), linear bearings (B), and shafts with support rails (C). ...............................................................................................35
22. Original adjustable mounting mechanism (A) rigidly fixed to the low-friction sled (B). The slotted holes (C) allowed for vertical movement. The clamp (D) allowed the leg holder shaft (E) to be adjusted in the direction tangent to the pendulum drop mass arc. ..........................................................................................36
23. Picture (left) and schematic (right) of the modified adjustable mounting mechanism. The original slotted holes (A) and clamp (B) remained. The additional set of slotted holes (C) allowed for transverse movement. ......................37
viii
8
24. Original stopping mechanism that allowed sled movement forward, but disallowed movement in the reverse direction. The teeth (A) of the stopping mechanism allowed for the ratcheting behavior. ......................................................38
25. Schematic (left) and picture (right) showing the locations of the sled lever (A) and cam follower (B). The schematic shows the lever position at rest, and the picture shows the lever position when it is in use. ....................................................39
26. Progression of the coil springs that were used for the pendulum device. The first spring is seen at the far left, progressing to the fourth and final spring on the far right. ...............................................................................................................41
27. Original data collection hardware and cart with wires extending to connect to the potentiometers and an available outlet. ...............................................................42
28. Final configuration of the data collection hardware fixed to the Delrin shelves. The top shelf contained the panel meter for easy viewing, and afforded ease of accessibility of the laptop computer. The bottom shelf contained the two power supplies (in the rear) that power each of the potentiometers, the battery pack for the foot trigger located (in the rear), the rotary switch on the left that cycled through each voltage output, two terminal blocks (in the front) for circuit organization, and the A/D converter (on the right). The foot trigger was placed at a desirable location on the floor by the user. ...........................................................................................................................44
29. Original smaller feet (2.54 centimeter diameter) with metal bottoms and the new larger feet (5.08 centimeter diameter) with rubber bottoms. The pictures on the left are for composition comparison. The picture on the right is for size comparison. ...............................................................................................................46
30. Three 25-pound (11.3-kilogram) weights attached to the rear of the pendulum device for added stability. Weight could be added or removed as needed. ..............47
31. Cross-supports (A) from the top left to the bottom right of the device. These additional supports provided a substantial increase in stability during impactions. ................................................................................................................47
32. Picture (left) and schematic (right) of the finalized pendulum device. The schematic shows the rotary potentiometer (A), pendulum arm (B), original pendulum drop mass (C), additional pendulum drop mass attached (D), sled components (E), laptop (F), panel meter (G), power supplies (H), rotary switch (I), terminal blocks (J), A/D converter (K), cross supports (L), power strip (M), battery (N) powering the foot trigger (O), and weights (P). The laptop and weights are not pictured on the left. ........................................................48
33. Close-up schematic of the finalized pendulum device. The locations of the rotary potentiometer (A), pendulum arm (B), original pendulum drop mass (C), additional pendulum drop mass attached (D), linear potentiometer (E), linear bearings (F), sled (G), adjustable mounting mechanism (H), coil spring (I), sled lever (J), and cam follower (K) are as noted. ..............................................49
34. Pendulum device with indenter rigidly fixed with the mounting mechanism and foam specimen sitting on the low-friction sled. .................................................53
ix
9
35. Post-impact foam specimens with visual indentations. Each specimen was impacted twice—once on the bottom and once on the top. The top six impactions were conducted at an initial drop angle of 20°, the middle six at 40°, and the bottom six at 60°. For some impactions at the low drop angle, two indentations were evident in a single impaction. At the higher drop angles, two indentations did not occur. The penetration depth clearly increased with an increase in drop angle. .................................................................55
36. Relationship between energy absorption calculations from the potentiometer data and MTS data. ...................................................................................................55
37. Motion capture system setup for the foam surrogate specimens (top) and cadaver leg porcine hock specimens (bottom). The markers and specimen deformation metric are labeled. ................................................................................58
38. Kinetic energy, sled energy, and specimen deformation histories for a representative foam impaction. .................................................................................61
39. Kinetic energy, sled energy, and specimen deformation histories for a representative cadaver leg porcine hock impaction. .................................................62
40. Pendulum device in vivo, including a rotary potentiometer that measured the pendulum’s linear velocity (A), the linear potentiometer that measured the sled’s linear displacement (B1), the specimen anchorage system (B2), the low-friction sled on linear bearings (B3), and the coil spring that resisted the sled’s forward motion (B4). ......................................................................................67
41. Orthopaedic surgeon checking for post-impact fracture. ..........................................68
42. Tibial plateau leveling osteotomy plate across the post-impact IAF insult. .............69
43. One-day post-fracture radiograph of the fracture insult and TPLO plate from Animal #1. ................................................................................................................71
44. Linear and rotary potentiometer voltage outputs for Animal #1. The linear displacement graph shows the sled beginning to translate at around 0.9 seconds (when impaction initiated). The displacement remained at 4.4 centimeters because the stopping mechanism was used. ..........................................73
45. Linear and rotary potentiometer voltage outputs for Animal #2. The linear displacement graph shows the sled beginning to translate at around 0.8 seconds (when impaction initiated). The displacement remained at 3.3 centimeters because the stopping mechanism was used. These voltage outputs were similar to those from Animals #3 and #4. ........................................................74
46. Linear and rotary potentiometer voltage outputs for Animal #6. The linear displacement graph shows the sled beginning to translate at around 0.9 seconds (when impaction initiated). The stopping mechanism was not used, so the linear displacement returned to zero. These voltage outputs were similar to those from Animals #7-11. .......................................................................77
x
10
A1. Circuit diagram of the data collection instrumentation and wiring. The letters correspond to the color of wire. The rotary switch has four paired nodes (nodes vertically opposite each other) that correspond to a position that could be cycled through and seen on the panel meter. .......................................................84
xi
11
LIST OF EQUATIONS
Equation
1. Equation and variables needed to calculate the initial gravitational potential energy of an impact mass. ...........................................................................................6
2. Equations and variables needed in the derivation and calculation of the pre-impact kinetic energy of the pendulum drop mass. ..................................................26
3. Equation and variables needed to calculate the post-impact sled energy. ................28
4. Equation and variables needed to calculate the energy absorption during fracture. .....................................................................................................................30
1
CHAPTER 1: INTRODUCTION
Osteoarthritis (OA), often referred to as the wear-and-tear arthritis, is the most
common joint disease [1]. Its symptoms include joint pain, stiffness, and limited range of
motion, which all lead to a lower quality of life [1-5]. Osteoarthritis can be divided into
primary and secondary types. Primary OA is an idiopathic phenomenon, while secondary
OA is caused by a number of risk factors including injury, obesity, inactivity, and
genetics [1-6].
Posttraumatic OA is a subset of secondary OA that develops after distinct
synovial joint injury. Synovial joint injuries such as meniscus tears, ligament tears,
capsule tears, joint dislocation, chondral injuries, and intraarticular fractures (IAFs) are
usually consequences from vehicle accidents, sports, falling, military injuries, or other
physical trauma sources [6-11]. To list a few examples of prevalence: lower-extremity
injuries from vehicle accidents are the main cause of automobile-related disabilities,
approximately 60%-90% of ruptured anterior cruciate ligaments result in posttraumatic
OA within 15 years, and over 12% of war-related trauma involves lower-extremity
injuries [12-15].
In addition to physical injury and joint degeneration, posttraumatic OA also has
significant financial burden [16-24]. It has been recently shown that $12 billion (12%) of
the overall annual cost of arthritis treatment in the United States is attributed to lower-
extremity posttraumatic OA (Figure 1). Approximately $3.06 billion (26%) of the annual
lower-extremity posttraumatic OA costs come directly from U.S. health care expenses.
This corresponds to 5.6 million Americans with posttraumatic OA symptoms that are
severe enough to require medical attention [7]. Furthermore, patients with posttraumatic
OA are frequently younger individuals who are in their prime income earning years [25-
27].
2
Figure 1. The cost of treatment of arthritis in the United States per year is $114.5 billion, and the distribution between rheumatoid arthritis (RA), non-posttraumatic OA, and posttraumatic OA is demonstrated. Image adapted and reproduced from Brown et al. [7].
It is evident that posttraumatic OA is of significant importance to society
physically and financially. Consequently, posttraumatic OA has been drawing increasing
attention. Despite progress in this research field during the last couple of decades,
however, the mechanisms leading to posttraumatic OA are still not fully understood.
Additionally, advancements in surgical treatment techniques have progressed, but the risk
of posttraumatic OA has still remained unacceptably high (Figure 2) [28]. In fact, over
half of patients with tibial plafond fractures develop posttraumatic OA within two years,
even with conventional treatment strategies [29-31].
In order to progress in this field, it is first necessary to create appropriate survival
animal models in which posttraumatic OA predictably develops in a manner consistent
with that in clinical settings. Once these models have been developed, they will enable
effective preclinical trials of prevention and treatment therapies. Such models need to
reproducibly replicate the pathophysiology involved in posttraumatic OA. There are
several well-recognized factors that contribute much to the pathogenesis of posttraumatic
OA [25, 32-35]. Acute factors include mechanical cartilage injury that involves physical
3
damage of collagen network and cartilage matrix, accompanied by chondrocyte death that
causes dysfunction of cartilage metabolism. Sub-acute factors include biological stimuli
leaked from damaged tissue, and those released associated with inflammatory reactions.
Chronic factors include cumulative abnormal mechanical stresses to damaged cartilage.
These factors are thought to contribute both individually and synergistically to the
development of cartilage degeneration.
Figure 2: Percentages of fair and poor outcomes of acetabular and tibial plateau intraarticular fracture treatments. Images adapted and reproduced from McKinley et al. [28].
Of these major factors, acute cartilage injury has been modeled by means of blunt
impaction [15, 36-45]. Unfortunately, previous survival animal models using this insult
modality did not predictably develop distinct cartilage degeneration in a short- to mid-
term [15, 37-39, 41, 42], presumably due to relatively mild cartilage damage, and/or due
to a lack of the chronic mechanical factor. Recently, Furman et al. [25] developed a
technique to model IAF in the mouse knee in vivo, and the insulted joints developed
massive cartilage degeneration over a period of 8 weeks. This is a very significant
experimental advance, although in its present embodiment it appears limited to relatively
4
severe fractures, and it exhibits substantial and uncontrolled variability in the location of
the fracture lines.
Regardless of insult modality, controllability and reproducibility of achievable
injury are crucial for this type of animal model to be effective. Unless consistency of
injury severity across experimental groups is ensured, preclinical trials of new treatment
cannot be scientifically valid. For this reason, use of an instrumented device capable of
measuring the properties of mechanical insult for each individual animal is highly
desirable.
1.1 Objective
Development of a large animal survival model of human IAF has been a long-
standing work in The University of Iowa Orthopaedic Biomechanics Laboratory. An
impaction insult technique to create IAFs in porcine hocks in vivo, with the injury
mechanisms consistent with those in human clinical injuries, had been developed. A
pendulum device that enabled implementing the fracture insult in survival surgery
settings also had been designed and built. As noted above, the next logical step was to
establish a technique to measure the properties of impaction insult that was executed to
create fracture in each individual animal. The purpose of the present work was to design,
construct, and evaluate a system to measure the energy absorbed during impaction
fracture insult.
In order to guide the initial direction of the proposed model, it was important to
investigate other blunt impaction and IAF models that have been developed. Out of these
many models, this work will discuss the blunt impaction models developed by Borrelli et
al. [36] and Tochigi et al. [46] and the IAF models developed by Furman et al. [25] and
Backus et al. [47]. These models were chosen because they have been prominent in
posttraumatic OA research, and because they incorporate a broad spectrum of ideas and
techniques that are commonly used among other blunt impaction and IAF models.
5
1.2 Blunt Impaction Models
1.2.1 Borrelli et al. Pendulum Device
Borrelli et al. [36] designed a pendulum device that delivered energy to the
femoral condyle in a rabbit model (Figure 3). The pendulum arm was 0.91 meters in
length and its mass was 587 grams. Additional mass could also be added to the pendulum
if desired. An electromagnet allowed the pendulum arm to be held at a desired height,
and released when necessary. When released, the pendulum arm impacted a padded low-
friction trolley equipped with a rigidly fixed impaction platen that transmitted a
compressive force pulse to the rabbit femoral condyle that was positioned using an X-Y
table. The platen was equipped with a load cell that collected force data at 2 kilohertz
using LabVIEW software.
With the dimensions and instrumentation described, it was possible to
conveniently acquire (i) the pendulum’s initial gravitational potential energy before
release, and (ii) the force history of impact. The initial gravitational potential energy,
which indicated the magnitude of energy input, was calculated with Equation 1 (below).
From the force history data, the maximum force and time to peak force could be acquired
as well. This is not unusual for devices equipped solely with load cells for
instrumentation [15, 40-44].
In a pilot study using cadaver specimens [36], pressure sensitive film was placed
between the impactor and the femoral condyle to measure contact area and contact stress.
These additional measurements were not applied to impaction insult in vivo, due to
difficulty in use of the pressure sensitive film in sterile surgical settings.
6
Figure 3: Close-up view of the components of the Borrelli et al. pendulum device. Pendulum arm (A), foam (B), trolley assembly (C), load cell (D), impactor (E), polyethylene block (F), and X-Y table (G). Image adapted and reproduced from Borrelli et al. [36].
Equation 1: Equation and variables needed to calculate the initial gravitational potential energy of an impact mass.
7
1.2.2 Tochigi et al. Drop Tower
Tochigi et al. [46] designed a drop tower device that delivers energy to the
femoral condyle in a rabbit model (Figure 4). The drop tower had a sliding drop mass
(1.55 kilograms) that could slide a maximum of 0.45 meters. Rather than attaching
additional mass like the Borrelli device, impaction magnitude could easily be controlled
by adjusting the drop height. A clamping release mechanism allowed the drop mass to be
held at a desired height and released when necessary. When released, the drop mass
impacted a platen that would then effectively impact the rabbit femoral condyle that was
positioned using a specimen fixation table. The drop mass was equipped with an
accelerometer. Acceleration data were collected at 15 kilohertz using LabVIEW software.
Figure 4: Tochigi et al. drop tower device in the surgical environment. The clamping release mechanism (A), the drop arm (B), and drop mass with accelerometer inside (C) are labeled.
8
With the dimensions and instrumentation described, it was possible to
conveniently acquire the drop mass’s pre-release potential energy (energy input) and the
acceleration history of impact. From the acceleration history data, the maximum
acceleration and time-to-peak acceleration could be ascertained as well. With the drop
mass being known, the acceleration data could be converted to force data.
Patterning after Borrelli’s work, a cadaver specimen impaction experiment using
pressure sensitive film was conducted to determine the contact area and contact stress of
achievable impaction. Again, these additional measurements were not applied to
impaction insult in vivo.
1.3 Intraarticular Fracture Models
1.3.1 Furman et al. Intraarticular Fracture Model
Furman et al. [25] designed a device that delivered energy to the tibial plateau in a
mouse model (Figure 5). The device utilized a materials testing system so that the user
could control the loading forces, rates of loading, and displacements of impaction.
Rigidly fixed to the materials testing system was a wedged-tip indenter that would deliver
a focalized compressive force pulse to the tibial plateau. The loading force and rate of
loading were 55 newtons and 20 newtons per second, respectively.
From the materials testing system, the force-displacement information could be
collected, and the corresponding energy of fracture could be calculated. When compared
to the devices with load cells and accelerometers, this technique clearly provides more
information. However, it is difficult to mimic a clinically-relevant impaction fracture
insult using a materials testing system, because trying to consistently control the high
load forces and rates of loading proves to be problematic. Evidence of this is present in
the results described by Furman et al. [25]. The energies of fracture that were recorded
exhibited a questionably large magnitude and range (33-231 joules). The quantity of the
9
data that is produced using a materials testing system is appealing, but the issue of
inconsistency can be difficult to overcome.
Figure 5: Cradle and indenter for creating an IAF in the mouse knee. Indenter and custom lower limb cradle for creating tibial articular fractures (left). Schematic showing alignment of indenter with tibia during loading (right). Image adapted and reproduced from Furman et al. [25].
1.3.2 Backus et al. Drop Tower Device
Backus et al. [47] designed a drop tower device to model IAF in the porcine knee,
by delivering a high-energy compressive force pulse across the joint (Figure 6). The drop
tower had a sliding drop mass (30 kilograms) that was released from a height of one
meter. When released, the drop mass impacted an aluminum specimen fixture, in which a
10
cadaver porcine knee (the whole-joint construct) was confined by holding the specimen’s
tibia and femur in place by using fiberglass reinforced resin and PMMA. The specimen
was preloaded (155 kilograms) using six parallel springs that were fixed to opposite ends
of the specimen fixture. Below the specimen fixture was a rigidly fixed load cell that
could record force history data.
With the dimensions and instrumentation described, it was possible to
conveniently acquire the drop mass’s pre-release potential energy (energy input) and the
force history of impact. It is important to note that, differently from the above-noted
systems, this impaction system was designed purely for fracture insult ex vitro, because
the potting and specimen fixture technique could not be used for in vivo scenarios.
Figure 6: Preload and impact alignment of closed porcine knee model. Posterior view of knee potted in aluminum fixtures and held with six springs (left). Dashed arrow represents indenter alignment to impact a knee without fracture; solid arrow represents the alignment to impact a knee resulting in an intraarticular fracture (right). Image adapted and reproduced from Backus et al. [47].
11
1.4 Limitations of Previous Models
Previous blunt impaction and IAF models have been fairly successful with
inducing posttraumatic OA, but there are still certain limitations that need to be addressed
before progressing further in model development. Some of the limitations were alluded to
in the previous section, but they will also be reaffirmed here. Models instrumented solely
with load cells are only able to collect force history information. Models instrumented
solely with accelerometers are only able to collect acceleration history information. With
a known drop mass, the acceleration history can be easily used to find force history
information. Therefore, models with load cells or accelerometers are limited to force
history data. The initial gravitational potential energy of impaction can be estimated
using the drop mass and drop height, but this is the only measurable type of energy in
these models. Models that use materials testing systems have attempted to provide energy
of fracture measurements from the force-displacement information, but the energies of
fracture are not consistent. In posttraumatic OA research, there has not been adequate
attention to acquiring fracture energy or energy absorption measurements. Most energy
measurements that are reported are the initial gravitational potential energy of impaction.
This is surprising, considering the variability between devices could cause inaccuracies in
the potential energy measurement. Friction, drag, and other factors could reduce the
impact velocity, meaning that not all of the potential energy was converted into kinetic
energy pre-impact. Additionally, it is naïve to say that the pre-impact energy is equal to
the fracture energy or energy absorbed. During impaction, pre-impact energy could be
transferred through a number of different pathways, e.g. specimen damage, soft tissue
deformation, vibrations, etc. Even though an accurate measure of fracture energy or
energy absorbed has not been focused on in posttraumatic OA research, however, it has
been used largely in other areas.
12
1.5 Energy Absorption Measurements
1.5.1 Charpy and Izod Impact Tests
Charpy and Izod impact tests were both developed in the early 1900s and are
currently American Society for Testing and Materials (ASTM) International methods for
determining energy absorption and impact strength [48]. Both tests use a pendulum
device that impacts and fractures a specimen. The essential concept of both tests is the
same: a pendulum with an initial pre-impact energy impacts and fractures a test
specimen, leaving the pendulum with a final post-impact energy (Figure 7).
Figure 7: Schematic illustrating the difference between the Charpy and Izod impact tests. Image adapted and reproduced from Callister [48].
13
The difference between pre- and post-impact energies is the energy absorbed by
the specimen during fracture. The specimens also have standardized dimensions and v-
notches for consistent fracture propagation lines. The primary difference between the
Charpy and Izod impact tests is the method in which the test specimens are held. The
Charpy impact test implements a three-point bending configuration, while the Izod
impact test implements a cantilevered beam configuration [48].
1.5.2 Abdel-Wahab et al. Izod Impact Test
Recent work by Abdel-Wahab et al. [49] showed the experimental results of an
Izod impact test of cortical bone tissue. Some cortical bone specimens were impacted at a
low energy level (0.02 joules), while others were impacted at a destructive energy level
(0.5 joules). The results showed that the energy absorption measurements were 20%
(0.004 joules) of the initial low energy level and 51% (0.256 joules) of the initial
destructive energy level. These results were confirmed numerically using extended finite
element methods. These results also compared favorably to the Charpy impact test results
published by Panagiotopoulos et al. [50]. It appears the Charpy and Izod impact tests are
effective for portions of bone tissue, but applying this to an IAF setting in vivo might be
challenging. Nevertheless, the overall basis of the Charpy and Izod impact tests does
confirm that it is naïve, and generally even incorrect, to say that the pre-impact energy is
equal to the energy absorbed.
1.5.3 Van Zeebroeck et al. Pendulum Device
Van Zeebroeck et al. [51] attempted to measure energy absorbed in impacts of
fruits and vegetables using an idea related to the Charpy and Izod impact tests. A
pendulum was designed and instrumented with a force sensor, accelerometer, and an
optical encoder (Figure 8).
14
Figure 8: Schematic representation of the pendulum arm with the optical encoder, force sensor, accelerometer, and impacted body (“fruit”) are indicated. Image adapted and reproduced from Zeebroeck et al. [51].
The optical encoder, located at the hinge of the pendulum arm, measured the
angular position of the pendulum. With the angular position known, the pre- and post-
impact kinetic energies could be calculated. As in the Charpy and Izod impact tests, the
difference between pre- and post-impact energies was the energy absorbed by the
specimen. The utilization of an optical encoder to experimentally calculate the kinetic
energy of the pendulum arm provides a more accurate representation of the pre- and post-
impact energies when compared to the theoretical potential energy calculations based on
15
drop height. The kinetic energy calculation is based on the actual pre- and post-impact
velocities, while the potential energy calculation takes no account of actually absorbed
energy. Furthermore, the optical encoder could ultimately be used to measure the energy
absorbed, since the pre- and post-impact kinetic energies were readily available. Even
though this particular application of the pendulum device was for impacting fruits and
vegetables, instrumenting a blunt impaction or IAF model with an optical encoder seems
to be advantageous.
1.6 Moving Forward
The information from the literature was important for establishing the initial
direction of the IAF model. After reviewing the models present in the literature, the
present work of designing, constructing, and evaluating an IAF model began.
16
CHAPTER 2: PRIOR DEVELOPMENT OF THE
INTRAARTICULAR FRACTURE MODEL
This chapter discusses the work that was conducted prior to the author’s presence.
This work established the foundation for the development of the IAF model.
2.1 Offset Impaction Technique
A drop tower fracture device was developed that used a compressive impaction
technique to produce distal tibial fractures in porcine cadaver legs. These distal tibial
fracture patterns were consistent with comparative human ankle fractures created by the
same drop tower fracture device (Figure 9). However, the severity of fracture was
deemed too severe for usage in future survival studies. In order to maintain a controllable
fracture pattern and to reduce the fracture severity for survival study applications,
alternative impaction techniques were pursued. A promising method of fracture was
found in an offset impaction technique [52]. With the offset impaction technique, the
tibia was rotated posteriorly with respect to the compressive force introduced by the drop
tower fracture device. This caused the compressive force to be primarily transmitted
through the anterior tibial juxta-articular bone. To ensure a consistent fracture pattern, a
small stress-rising saw cut (Figure 10) was applied using an oscillating bone saw. The
stress-rising saw cut was initiated at the anterior distal tibial surface and proceeded until a
thickness of approximately 5-7 millimeters of cancellous bone underlying the articular
surface remained intact. While not affecting the articular surface and associated
chondrocyte viability, the initial stress-rising saw cut at the anterior distal tibial surface
effectively ensured similar fracture line location and orientation after impaction, in a
preliminary cadaveric study [52]. Differences in fracture severity from two distinct
magnitudes of energy delivery could be seen, yet the fracture patterns still remained
consistent (Figure 10). The pattern of chondrocyte damage within these fractures was
shown to be analogous to the damage present in the previous human ankle fractures [52,
17
53]. With the offset impaction technique proving its efficacy in cadaveric studies, it was
necessary to alter the specimen mounting procedure to allow for in vivo applications.
Consequently, much effort and experimentation was consumed in developing a tripod
anchorage system that could successfully utilize the discussed offset impaction technique
in cadaveric and survival studies.
Figure 9: Drop tower fracture device (A), close-up of a mounted ankle specimen (B), three human ankle fracture patterns (C), schematic of porcine distal tibial fracture technique (D), and a single porcine fracture pattern (E). The fracture patterns between the human and porcine cases were comparable. Note the severity of fracture in the porcine fracture. Images adapted and reproduced from Tochigi et al. [53].
18
Figure 10: Schematic of the offset impaction insult (A), the introduced stress-rising saw cut technique (B), three fracture patterns from moderate energy impacts (40 joules) (C), and three fracture patterns from high energy impacts (60 joules) (D). The fracture patterns between the compression and offset impactions were comparable. By comparison with Figure 9, note that the severity of fracture in the offset impactions was less than in the compression impactions.
2.2 Tripod Anchorage System
A schematic of the tripod anchorage system is shown in Figure 11. The distal
tripod pins caused the impact force to be directly transmitted to the talus, which
minimized any unwanted energy dissipation via soft tissue during impaction. The distal
tibial plate and leg holder shaft maintained the position of the porcine tibia. The distal
tibial plate was secured with two pegs and by a cortical screw on the anterior distal tibial
surface. The plate contained a ball-in-socket joint that allowed the leg holder shaft to be
appropriately aligned with respect to the impact force. The leg holder contained an
external fixator pin that was fixed to the proximal tibia, which effectively maintained the
position of the tibia. This was important since the offset impaction technique created a
19
moment during impaction that would tend to rotate the position of the tibia if the tibia
was not securely attached to the leg holder.
Figure 11: Schematic (left) and radiograph (right) of the tripod device-to-bone anchorage system. Images adapted and reproduced from Tochigi et al. [52].
Preliminary studies with the drop tower fracture device had tested the tripod
anchorage system, using similar techniques as before. The fracture patterns with the
tripod anchorage system proved to be comparable to the fracture patterns within the
cadaver leg study discussed in the previous section (Figure 12). In addition, the pattern of
chondrocyte damage and its time-dependent progression agreed with earlier data from
human ankle fractures and porcine cadaver leg study discussed in the previous section
(Figure 12). Chondrocyte damage was further investigated using a surgical osteotome as
a non-impact osteotomy control to compare results. It was found that for osteotome
controls, fractional cell death was significantly lower in the near-edge regions when
compared to the impaction fracture joints (Figure 13) [53]. This showed that impaction-
20
fractured joints experienced greater mechanical stresses than the non-impact osteotomy
control joints. Collectively, the preliminary studies that had been performed showed that
the offset impaction technique along with the tripod anchorage system was capable of
mimicking clinically realistic acute mechanical cartilage damage, in a way that traditional
non-impact fracture techniques could not accomplish. Furthermore, the offset impaction
technique using the tripod anchorage system appeared to be appropriate for introducing
pathophysiologically realistic cartilage injury in future cadaveric and survival studies.
However, the drop tower fracture device did not allow live animals to be suitably
positioned for anesthesia and surgical management in survival studies. To this end, a
pendulum device was developed that incorporated the same offset impaction technique
with the tripod anchorage system.
Figure 12: Three fractures created using the tripod anchorage system in the porcine tibia (left). Note the general consistency of fracture line location. Representative time-tracking confocal microscope images at a near-fracture site produced using the tripod anchorage system in the porcine tibia (middle) and using the compressive impaction technique in the human ankle (right). Cells labeled green were alive, while red cells were dead. Arrows indicate the edge of the fragment. Image on right adapted and reproduced from Tochigi et al. [53].
21
Figure 13: Chondrocyte viability results (2 days post-impact) in impaction-fractured joints vs. non-impact osteotomy controls. Dispersion bars indicate 75th- and 25th-percentile values.
2.3 Original Pendulum Device
The original Orthopaedic Biomechanics Laboratory pendulum device is shown in
Figure 14. It was built using Computed Numerically Controlled (CNC) machines (a
HAAS Tool Room Mill 1 and a HAAS Precision Collet Lathe). The structure was
constructed primarily with 6061 aluminum alloy. The pendulum drop mass (3.88
kilograms) was made of 303 austenitic stainless steel alloy.
The energy of impaction could be controlled by adjusting the drop angle (height)
of the pendulum arm. A release mechanism was also utilized to consistently release the
pendulum arm for impaction. The adjustable support beams allowed the height of the
specimen table and pendulum drop mass to be raised or lowered to the desired location. A
specimen was positioned according to the offset impaction technique and fixed to the
table using the bone-to-device tripod anchorage system.
22
Figure 14: Original Orthopaedic Biomechanics Laboratory pendulum device. The pendulum arm (A), release mechanism (B), support beams (C), specimen table (D), and pendulum drop mass (E) are shown.
2.4 First Application of the Original Pendulum Device
In a pilot study, the pendulum device performed well in creating desirable fracture
patterns and inducing progressive chondrocyte damage near the fracture region (Figure
15). These results were comparable with the earlier preliminary studies that had used the
drop tower fracture device. Therefore, the pendulum device seemed to be an attractive
method to create pathophysiologically realistic cartilage injury in future cadaveric and
survival animal studies.
With the general concept established, there was a logical need for quantification.
The original Orthopaedic Biomechanics Laboratory pendulum device did not have any
type of instrumentation, so the only impaction magnitude value that could be quantified
was the pendulum’s initial gravitational potential energy (before release). This situation
was not a suitable basis for future work, so appropriate instrumentation had to be pursued
to accurately measure the amount of energy absorbed by a specimen during fracture
23
insult. Appropriate modifications and additions necessary to accommodate
instrumentation of the pendulum fracture device are discussed in the next section.
Figure 15: Porcine distal tibia fractures created using the pendulum impaction system.
24
CHAPTER 3: DEVELOPMENT OF THE PENDULUM
DEVICE AND DATA COLLECTION TECHNIQUE
This chapter discusses the author’s contribution to the design of the pendulum
device and its instrumentation for collecting energy absorption measurements. Details
concerning the appropriate modifications, additional components, and software
techniques for ease of data extraction are provided.
3.1 Initial Pendulum Instrumentation, Modifications, and
Additions
3.1.1 Overview
As with Charpy and Izod impact tests, the energy absorbed by a specimen during
fracture was calculated as the difference between pre- and post-impact energy
measurements. The instrumentation, modifications, and additions to the pendulum device
were centered around this idea. If the pre- and post-impact energies could be accurately
measured, then the energy absorption value could be deduced.
Figure 16 shows the general components of the instrumented pendulum device. A
brief overview is discussed in this section, while more detail is included in the following
sections. The rotary potentiometer measured the angular position of the pendulum arm.
The pendulum drop mass was released from a desired drop height. The initial
gravitational potential energy was converted into kinetic energy as the drop mass
accelerated downward. The pre-impact kinetic energy could be calculated from the rotary
potentiometer information. The drop mass impacted the specimen that was rigidly fixed
to the low-friction sled on linear bearings. The linear displacement of the sled was
calculated from the linear potentiometer, which measured the linear position of the sled.
The sled’s forward motion was resisted by a coil spring. The stopping mechanism
prevented the sled from recoiling after impaction. The sled lever acted as a second-class
25
lever with the cam follower as the resistive loading point. The post-impact energy could
be calculated from the linear potentiometer information and the spring constant.
Figure 16: Picture of the general components of the instrumented pendulum device. The angular position (Angle) of the pendulum arm measured by the rotary potentiometer and the linear displacement (x) measured by the linear potentiometer are shown.
3.1.2 Rotary Potentiometer
As illustrated by the methodology of Zeebroeck et al. [51], the kinetic energy of a
pendulum at the instant just before impact (pre-impact energy) can be measured with an
optical encoder located at the hinge of the pendulum arm. This type of measurement
technique was attractive for instrumenting the original Orthopaedic Biomechanics
Laboratory pendulum device. Ultimately, a rotary potentiometer was attached at the
hinge of the original pendulum device. A rotary potentiometer is a three-terminal resistor
26
with a sliding contact that acts as an adjustable voltage divider. With a given power
supply, the voltage output of a rotary potentiometer can range from zero to the given
power supply voltage, depending upon the position of the sliding contact. This sliding
contact is a rotational sliding contact. Two of the three leads of a rotary potentiometer are
for the power supply, while the third lead is for the voltage output. For the voltage output
to be meaningful, the rotary potentiometer had to be calibrated. Using a 10-volt power
supply, the voltage output of the rotary potentiometer was recorded in 20° increments
ranging from 0° to 240°. From this information a calibration curve was fit to obtain a
relationship between the voltage output and the angular position of the pendulum arm.
With the properly calibrated rotary potentiometer and known pendulum arm length, the
angular position of the pendulum arm could be converted into linear position information
by implementing basic trigonometry. Since velocity is the derivative of position, the
linear velocity of the pendulum drop mass could be computed by finding the slope of the
real-time linear position information. The linear velocity could then be used to calculate
the kinetic energy of the known pendulum drop mass. The mathematical derivation and
corresponding units of the described method are outlined in Equation 2.
Then, the rotary potentiometer could be used to effectively find the kinetic energy
of the pendulum drop mass at any time point. Therefore, the pre-impact kinetic energy of
the pendulum drop mass could be measured. Given this technique of quantifying the pre-
impact energy, a means for quantifying the post-impact energy was also necessary to
deduce an energy absorption value. In order to achieve post-impact energy measurement,
a sled system was designed and built.
27
Equation 2: Equations and variables needed in the derivation and calculation of the pre-impact kinetic energy of the pendulum drop mass.
3.1.3 Sled System
The sled system consisted of a low-friction sled on linear bearings, a linear
potentiometer, a coil spring, an adjustable mounting mechanism, and a sled lever. Figure
16 provides a visual reference that shows the locations of the sled system components.
The low-friction sled sits on linear bearings that allow the sled to move in the same
direction as the pendulum drop mass, while restricting motion in the other two directions
(vertical and transverse).
The linear potentiometer was attached at the base of the low-friction sled. The
linear potentiometer works essentially in the same manner as the rotary potentiometer.
28
Both are three-terminal resistors with a sliding contact that acts as an adjustable voltage
divider. The sliding contact for a linear potentiometer is straight-line, while a rotary
potentiometer has a rotational sliding contact. Similar to the rotary potentiometer, the
linear potentiometer had to be calibrated in order for the voltage output to be meaningful.
Using a 10-volt power supply, the voltage output of the linear potentiometer was
recorded in 5-millimeter increments ranging from 0 to 50 millimeters. From this
information a calibration curve was fit to obtain a relationship between the voltage output
and the linear position of the low-friction sled. With the properly calibrated linear
potentiometer, the linear position of the low-friction sled could be found at any time
point.
The linear position information alone could not determine the post-impact energy,
so a coil spring was introduced. The coil spring was positioned such that the forward
motion of the low-friction sled was resisted after impaction. Again, the coil spring had to
be properly calibrated to provide meaningful information. Using an MTS 810 materials
testing machine (MTS Systems Corp., Eden Prairie, MN), the force required to compress
the spring at controlled displacements was recorded in 5-millimeter increments ranging
from 0 to 50 millimeters. From this information a calibration curve was fit to obtain the
spring constant. With the spring and linear potentiometer, the acquired spring constant
and the linear position information of the low-friction sled could be used to calculate the
spring energy required to stop the low-friction sled’s forward progress (post-impact
energy). This energy is the potential energy that is stored in the spring when it is
compressed by the low-friction sled. To avoid confusion with the initial gravitational
potential energy of impaction, the energy described will be referred to as “sled energy”
for the remainder of this work. The equation used to calculate the sled energy from the
spring constant and the linear position information of the low-friction sled is outlined in
Equation 3. The linear displacement of the low-friction sled was equivalent to the
compressive length of the spring, because the spring was initially at its equilibrium
29
position (not prestressed), and the spring and the low-friction sled were in intimate
contact.
Equation 3: Equation and variables needed to calculate the post-impact sled energy.
The sled system also needed a method for attaching the tripod anchorage system
used in the offset impaction technique for creating IAFs. To this end, an adjustable
mounting mechanism was rigidly fixed to the low-friction sled. The mounting mechanism
allowed the tripod anchorage system to be adjusted in the direction of the pendulum drop
mass, as well as the vertical direction. Adjustability in the transverse direction was not
allowed, since it was deemed unnecessary.
Due to the intrinsic behavior of a spring, there would be recoil after release of the
compressing force compressed. In an attempt to minimize this effect, a stopping
mechanism was designed and built. The stopping mechanism was a simple ratchet
mechanism that allowed movement in the direction of impact or compression, but that
halted any movement in the opposite direction (direction of recoil).
Since the stopping mechanism would need to be released to return the spring to its
equilibrium position, a sled lever was attached for assistance. A cam follower was
30
attached to the low-friction sled to provide a point of contact between the sled and sled
lever. The dimensions of the sled lever, along with the location of the cam follower and
sled lever, generated a mechanical advantage of 6. This was more than enough for easy
operation of the sled lever.
3.1.4 Data Collection Components
The analog voltage outputs of the potentiometers needed to be digitized and
collected on a computer to calculate the energy measurements. A power supply, analog-
to-digital (A/D) converter, and computer had to be available to collect impaction data
from the potentiometers. For initial trials with the pendulum device, a dual-output power
supply (Model #E3620A, Agilent Technologies Inc., Santa Clara, CA) was used to
supply 10 volts to each of the two potentiometers. The dual-output power supply also had
a panel meter that displayed the output voltage of a selected potentiometer. An NI USB-
6210 (National Instruments Corp., Austin, TX) A/D converter was used with a Dell XPS
laptop (Dell Inc., Round Rock, TX) to collect the voltage output readings from the
potentiometers. To minimize the amount of unwanted data, a 3-volt AC powered hand
trigger was used. This particular trigger voltage was chosen because 3 volts is an
effective voltage for digital triggers—digital triggers require a transistor-to-transistor
logic (TTL) voltage level of 1-5 volts.
The pendulum device and the additional components could collect position data
from the rotary and linear potentiometers. From these data, the pre-impact kinetic energy
of the pendulum drop mass and the post-impact sled energy could be calculated. The
energy absorbed by the specimen during impaction could then be calculated as the
difference between the pre-impact kinetic energy and the post-impact sled energy
(Equation 4).
31
Equation 4: Equation and variables needed to calculate the energy absorption during fracture.
3.2 Final Pendulum Instrumentation, Modifications, and
Additions
Once the initial pendulum instrumentation, modifications, and additions were
completed, many preliminary impaction trials were conducted to find aspects that needed
improvement. The preliminary impaction trials and the variables that were tested can be
seen in Table A1 in the Appendix. This section discusses the development phase for each
major component. Topics covered include preliminary impaction results, failure
modalities, modifications, reasoning for modifications, and fine-tuning.
3.2.1 Rotary Potentiometer
Figure 17 shows the rotary potentiometer attached at the hinge of the pendulum
arm. After preliminary impactions were conducted, the rotary potentiometer proved to be
quite sufficient. The signal from the rotary potentiometer was relatively noise-free, so
calculating the slope of the linear position information was not problematic. Based on the
information from the rotary potentiometer, it was clear that the linear velocity of the
pendulum monotonically increased until the event of impaction, such that the maximum
linear velocity occurred at the point of incipient impaction. Logically, this would be
expected to happen, because the pendulum drop mass was accelerating during the
32
downward swinging motion. Since the maximum linear velocity could be found with
relative ease, the pre-impact kinetic energy was also easily calculated. Based on the
initial impaction results, the pre-impact kinetic energy was approximately 93% of the
pendulum’s pre-release gravitational potential energy. This was reasonable, in that the
pre-impact kinetic energy was expected to be less than the initial gravitational potential
energy due to friction, drag, and other factors that would decrease the linear velocity of
the pendulum drop mass. Since the signal from the rotary potentiometer was relatively
easy to work with and since the magnitude of the pre-impact kinetic energy
measurements seemed to be very reasonable, the rotary potentiometer was deemed to
need no further modifications. The original rotary potentiometer was therefore used
throughout future impactions.
Figure 17: Side view (left) and front view (right) of the rotary potentiometer attached to the hinge of the pendulum arm.
3.2.2 Linear Potentiometer
Figure 18 shows the original linear potentiometer that was attached to the low-
friction sled. Unlike its rotary counterpart, the linear potentiometer caused concerns even
before preliminary impactions. The sliding contact had a significant amount of drag in
33
some positions of its traverse. In addition to the drag issue, the original linear
potentiometer only accommodated 6.1 centimeters of traverse. Depending on the
impaction magnitude, this amount of travel sometimes exceeded this distance, causing
the linear potentiometer to be bottomed out. Bottoming out a potentiometer obviously is
undesirable, for multiple reasons. To start with, if the maximum travel were reached
during an impaction, the sled energy information would be incorrect. The measured
compression distance would be too low, leading to a subsequent under-report of post-
impact sled energy. In addition to measurement inaccuracies, bottoming out could have
damaging effects to the potentiometer itself.
Figure 18: Original linear potentiometer that was attached to the low-friction sled.
Therefore, a new linear potentiometer was investigated (Figure 19). The drag
from the new sliding contact was insignificant when compared to that from original linear
potentiometer. This resulted in smoothly accurate linear displacement readings. The new
linear potentiometer also allowed for a linear travel up to 15.2 centimeters, almost a
150% increase when compared to the original linear potentiometer. This would prevent
34
any of the problems associated with bottoming out the potentiometer. To calibrate the
voltage output signal of the new linear potentiometer, again a 10-volt power supply was
used. The voltage output was recorded in 5-millimeter increments ranging from 0 to 65
millimeters. Again, this information was used to fit a calibration curve and obtain a
relationship between the voltage output and the linear position of the low-friction sled.
Once the new linear potentiometer was calibrated and properly attached to the low-
friction sled, preliminary impactions were conducted. The signal from the linear
potentiometer was still not as “clean” as that from the rotary potentiometer. However, the
magnitude of the noise was not large enough to cause problems with measuring the linear
displacement. Plus, simple signal processing techniques could always be applied to filter
out the unwanted noise frequencies. With the drag being insignificant, the travel being
more than sufficient, and the signal being relatively clean, the new linear potentiometer
was deemed adequate for use in future impactions, and no further modifications had to be
made.
Figure 19: New linear potentiometer that was attached to the low-friction sled.
35
3.2.3 Low-Friction Sled and Linear Bearings
Figure 20 shows the original low-friction sled and linear bearings. The original
low-friction sled and linear bearings were part of another device, so they were used
temporarily as proof of concept. Nevertheless, its temporary use contributed to the design
process of the definitive low-friction sled and linear bearings. Initial observations
revealed that minor surface defects (corroded shaft with support rail) contributed to a
slight increase in friction that could affect the post-impact sled energy measurements.
Figure 20: Schematic of the original low-friction sled (A), linear bearings (B), and shafts with support rails (C).
Trial impactions showed that a low-friction sled and linear bearings could in
principle be effectively used for the desired application. Post-impact sled energies could
be gathered, and their magnitudes seemed reasonable. No insurmountable difficulties
arose during trial impactions that indicated need to abandon the essential concept. With
the proof-of-concept established, a new low-friction sled and linear bearings needed to be
procured to replace the temporary version.
36
Figure 21 shows the new low-friction sled and linear bearings. The sled table was
built from 6061 aluminum alloy using a HAAS Tool Room Mill 1. Linear sleeve open
bearings (Part #6374K314, McMaster-Carr, Elmhurst, IL) were purchased and fixed to
the underside of the sled table and aligned with a purchased shaft with support rail (Part
#6557K22, McMaster-Carr, Elmhurst, IL).
Preliminary impactions with the new low-friction sled and linear bearings
exhibited slight improvement from its original counterpart. Post-impact sled energies
were gathered, and their magnitudes were similar to those of the original unit. Through
manual observation, the frictional component of the new bearings was less than that of
the previous bearings. This was probably attributed to the pristine condition of the new
components, specifically the non-corroded surface of the shaft with support rail. From the
preliminary impactions and observations, the new low-friction sled and linear bearings
could successfully be used in future impactions.
Figure 21: Schematic of the new low-friction sled (A), linear bearings (B), and shafts with support rails (C).
37
3.2.4 Adjustable Mounting Mechanism
The original adjustable mounting mechanism can be seen in Figure 22. The
mounting mechanism had slotted holes that allowed the tripod anchorage system to be
adjusted in the direction tangent to the pendulum drop mass arc, as well as in the vertical
direction. When the original mounting mechanism was applied to in vivo cases, more
adjustability was found to be needed for most animals. The adjustability in the direction
of the pendulum drop mass was more than sufficient, but more adjustability was often
needed in the downward vertical direction. Due to it being coupled to the overall body of
the animal, the hock could not be rotated with ease, unlike the case for the cadaveric leg
(limb-only) impactions. This caused the need for adjustability in the transverse direction.
Figure 22: Original adjustable mounting mechanism (A) rigidly fixed to the low-friction sled (B). The slotted holes (C) allowed for vertical movement. The clamp (D) allowed the leg holder shaft (E) to be adjusted in the direction tangent to the pendulum drop mass arc.
To allow for the added adjustability, the original mounting mechanism was
modified (Figure 23). The modified mounting mechanism trimmed off excess material
that was limiting the full range of adjustability in the downward direction. These
38
modifications allowed the tripod anchorage system to be adjusted downward an
additional 2.5 centimeters. The modified mounting mechanism also included a different
plate, with slotted holes that allowed for adjustability transversely approximately ±1.9
centimeters. Following additional in vivo trials, excess material was trimmed off the
newly-added plate. This final modification was made because the new plate would
occasionally make contact with particularly large animals. The final modification did not
pose the same problem in future in vivo cases, even with the overweight specimens.
Figure 23: Picture (left) and schematic (right) of the modified adjustable mounting mechanism. The original slotted holes (A) and clamp (B) remained. The additional set of slotted holes (C) allowed for transverse movement.
3.2.5 Stopping Mechanism
The stopping mechanism allowed movement in the direction of impact or spring
compression, but halted any movement in the opposite direction. The original stopping
mechanism is shown in Figure 24. Initially, it was felt to be desirable to minimize the
recoil of the spring, in an attempt to prevent additional damage to the specimen.
However, after analyzing high-speed video data (Oqus Camera Series, Qualisys AB,
39
Gothenburg, Sweden), the recoil of the spring was found not to be aggressive as
presumed. In fact, the recoil was quite smooth, gentle, and non-destructive to the
specimen. Given this additional information, the idea of the stopping mechanism was
reevaluated. In preliminary impactions, the original stopping mechanism was not always
successful in halting movement in the opposing direction. Sometimes the stopping
mechanism would bounce back in a skipping manner, which caused a jerking motion of
the specimen. Even when the stopping mechanism was successful, other concerns
emerged. Due to the ratchet nature of the stopping mechanism, it inherently produced
friction during forward movement. This would cause the linear displacement of the low-
friction sled to decrease, and the subsequent post-impact sled energy to be erroneously
low. In addition, the stopping mechanism protruded in such a fashion there was contact
with the specimens in vivo. That contact led to substantially lower linear displacement
and thus post-impact sled energy measurements. From the observations of its
shortcomings, the stopping mechanism did not prove to be advantageous in any way.
Although the stopping mechanism seemed beneficial in theory, in practice it did not
perform as planned. Ultimately, this component was removed from the pendulum device,
in the interest of making more accurate measurements.
Figure 24: Original stopping mechanism that allowed sled movement forward, but disallowed movement in the reverse direction. The teeth (A) of the stopping mechanism allowed for the ratcheting behavior.
40
3.2.6 Sled Lever
The location of the sled lever with respect to the low-friction sled and the cam
follower can be seen in Figure 25. With the stopping mechanism removed, the original
purpose of the sled lever was not applicable. However, secondary functions of the sled
lever arose during preliminary tests. While setting up the pendulum device for data
collection, it was desirable to test the potentiometers first to ensure accuracy. In order to
test the linear potentiometer, the coil spring had to be compressed for the low-friction
sled to move in the direction of the pendulum drop mass. Compressing the spring by hand
was possible, but extremely difficult when the compression displacement was large.
When the sled lever was attached, it allowed the user to maneuver the sled and coil spring
with relative ease. Even though the original purpose of the sled lever was not applicable
with the stopping mechanism removed, the secondary function of enabling testing of the
linear potentiometer prior to impaction was still important. Therefore, the sled lever and
cam follower remained in their original positions for future impactions.
Figure 25: Schematic (left) and picture (right) showing the locations of the sled lever (A) and cam follower (B). The schematic shows the lever position at rest, and the picture shows the lever position when it is in use.
41
3.2.7 Coil Spring
Figure 26 shows the progression of coil springs that were used for the pendulum
device. The respective spring constants, maximum compression displacements, and total
energy capacities are shown in Table 1. The first spring was chosen for low-energy
impaction tests initially. After it became apparent that the energy capacity of the first
spring was too small for fracture, it was replaced with the second spring. The second
spring performed well when impacting surrogate specimens, but preliminary impactions
with cadaver leg porcine hock specimens and the first in vivo specimen raised concern.
The compression displacement limits were neared frequently, so the energy capacity of
the second spring was being reached. This would cause inaccurate post-impact sled
energy readings and could damage the spring. So again, the spring had to be replaced.
The third spring performed very well in preliminary impactions with cadaver leg porcine
hock specimens. When transitioning to in vivo trials, the third spring initially showed
promise. However, its energy capacity started being approached when the stopping
mechanism was removed. From all the information that was gathered, it was decided that
the next spring should err on the “safe side” for possible energy capacity. To that end, the
fourth spring allowed for an energy capacity that neared the amount of pre-impact kinetic
energy. This was felt to be sufficient to allow for any plausible changes that might have
been made to the device or surgical procedure. The larger energy capacity also allowed
for the possibility of impacting specimens other than porcine hocks in the future, perhaps
requiring higher energy capacity values. From in vivo impactions with the fourth spring,
the results were consistent and did not approach the spring’s energy capacity. The
preliminary results were very encouraging and showed that the fourth spring could
effectively be used for future impactions.
42
Figure 26: Progression of the coil springs that were used for the pendulum device. The first spring is seen at the far left, progressing to the fourth and final spring on the far right.
Table 1: The characteristics for the progression of coil springs, including the respective spring constants, maximum compression displacements, and total energy capacities.
3.2.8 Data Collection Hardware
The original data collection hardware and the cart that was used to house that
hardware can be seen in Figure 27. This configuration required many cords to connect
from the pendulum device, to the cart, and finally to an outlet. When maneuvering the
pendulum device and hardware cart during preliminary in vivo impactions, it was difficult
43
to protect the cords. Attention had to be paid so that the hardware cart would not run over
the cords. If the cart and pendulum device needed to be appreciably separated from each
other, the cords would be stretched. Overall, the configuration led to awkward and
stressful mobility issues. To keep the hardware more organized and connected, it was
desirable for the hardware to be attached to the pendulum device in some manner. To
accommodate for such attachment, two shelves were constructed from Delrin.
Figure 27: Original data collection hardware and cart with wires extending to connect to the potentiometers and an available outlet.
Also, the previous dual-output power supply could only be used temporarily,
because it was needed for other purposes. The new hardware that was purchased to
replace the previous dual-output power supply consisted of two 10-volt power supplies
(Part #PSS-10, Omega Engineering Inc., Stamford, CT), a panel meter (Part #DP41-E,
Omega Engineering Inc., Stamford, CT), and a rotary selector switch (Part #65865K32,
McMaster-Carr, Elmhurst, IL). The two 10-volt power supplies had the same
functionality as the original dual-output power supply in terms of providing the
44
potentiometers with 10-volt excitation voltages. The panel meter also had the same
functionality as that in the original dual-output power supply, in displaying the voltage
outputs of the potentiometers. When compared to the original dual-output power supply,
the rotary switch was an improvement for purposes of checking the voltage drops across
different components. With the original dual-output power supply, two electrical lead
hook clips had to be placed properly across a single component to view its corresponding
voltage drop. The hook clips had to be repositioned each time a different component was
of interest. When replaced with the rotary switch, each component’s voltage drop could
be viewed by simply changing the position of the rotary switch. Components to be
checked were the two power supplies and the two potentiometers. Before each impaction,
the voltage drops across the 10-volt power supplies were confirmed. Additionally, the
voltage drop across the linear potentiometer was checked to ensure that no inaccuracies
arose. Finally, the voltage drop across the rotary potentiometer was of primary interest.
The initial vertical position had to be confirmed for proper computation of the pre-impact
kinetic energy of the pendulum drop mass. It was also required to view the voltage output
of the rotary potentiometer, in order to set the appropriate drop angle (energy input) of
impaction. With the rotary switch attached, cycling through each desirable component
was much easier and more intuitive than using the previous dual-output power supply.
The two power supplies, panel meter, and rotary switch were all wired and fixed to the
Delrin shelves once their functionalities were confirmed in preliminary impactions.
In the same preliminary impactions, the original A/D converter and laptop
performed well in collecting the voltage output readings from the potentiometers. There
was no need for alterations, so they were also fixed to the Delrin shelves. The 3-volt AC
powered hand trigger that was used worked well, but required an additional operator. In
an attempt to keep the required operator numbers to a minimum, a 3-volt foot trigger
replaced the hand trigger. This allowed the same operator to trigger the data collection
and release the pendulum drop mass, effectively reducing the amount of required
45
operators to one. The foot trigger was also powered with two AA batteries to eliminate
the need for an AC power connection. With successful implementation of the foot trigger,
it remained on the pendulum device for future impactions. The final configuration of the
data collection hardware fixed to the Delrin shelves of the pendulum device can be seen
in Figure 28. The corresponding circuit diagram can be seen in Figure A1 in the
Appendix.
Figure 28: Final configuration of the data collection hardware fixed to the Delrin shelves. The top shelf contained the panel meter for easy viewing, and afforded ease of accessibility of the laptop computer. The bottom shelf contained the two power supplies (in the rear) that power each of the potentiometers, the battery pack for the foot trigger located (in the rear), the rotary switch on the left that cycled through each voltage output, two terminal blocks (in the front) for circuit organization, and the A/D converter (on the right). The foot trigger was placed at a desirable location on the floor by the user.
46
3.2.9 Pendulum Arm Modifications
The original pendulum arm had a thickness of 4.8 millimeters, which resulted in
undesired transverse movement that could cause an indirect impact. To alleviate this
problem, the pendulum arm thickness was increased to 9.5 millimeters, resulting in a
700% increase in the area moment of inertia when compared to the original arm
thickness. Preliminary impactions further reaffirmed that the tendency of the pendulum
arm to move transversely was greatly decreased with the thicker pendulum arm.
Preliminary impactions also indicated the need for a greater pendulum drop mass.
The original pendulum drop mass (3.88 kilograms) only allowed for a maximum pre-
impact energy of approximately 38 joules at a drop angle of 90°. Larger pre-impact
energies were found to be required for fractures in vivo, so the pendulum drop mass was
increased to 5.82 kilograms. The larger pendulum drop mass allowed for a maximum pre-
impact energy of approximately 57 joules at a drop angle of 90°, an effective increase of
50%. In preliminary in vivo fractures, the larger pendulum drop mass was successful in
consistently creating fractures, while the smaller proved insufficient for fracture.
3.2.10 Stability Modifications
Finally, modifications were made that improved the stability of the pendulum
device during impaction. Preliminary in vivo trials showed that the overall stability of the
pendulum device was not sufficient for controlled impactions. Noticeable problems
included the entire pendulum device sliding forward in the direction of the pendulum
drop mass, tilting up on its front two feet, and bowing forward during impaction. To
prevent the pendulum device from sliding forward on the floor, the feet were replaced
with larger ones that were made of rubber (Figure 29). The larger surface area and usage
of rubber material provided more friction on the floor, and successfully prevented any
forward sliding in the direction of the pendulum drop mass. To prevent the pendulum
device from tilting up on its front two feet, three 25-pound (11.3-kilogram) weights were
47
attached to the rear (Figure 30). This solved the pendulum tilting problem in preliminary
impactions, and of course more weight could always be added if needed. To prevent the
pendulum device from bowing forward, cross supports were attached to the rear as well
Figure 31. This modification was the main source of improved structural stability. With
these stability modifications in place and tested, the pendulum device was finally ready to
be put to use in actual in vivo impaction studies. The final pendulum configuration is
shown in Figure 32 and Figure 33. Before describing the actual impaction studies and
results, however, the data collection technique needs to be set forth.
Figure 29: Original smaller feet (2.54 centimeter diameter) with metal bottoms and the new larger feet (5.08 centimeter diameter) with rubber bottoms. The pictures on the left are for composition comparison. The picture on the right is for size comparison.
48
Figure 30: Three 25-pound (11.3-kilogram) weights attached to the rear of the pendulum device for added stability. Weight could be added or removed as needed.
Figure 31: Cross-supports (A) from the top left to the bottom right of the device. These additional supports provided a substantial increase in stability during impactions.
49
Figure 32: Picture (left) and schematic (right) of the finalized pendulum device. The schematic shows the rotary potentiometer (A), pendulum arm (B), original pendulum drop mass (C), additional pendulum drop mass attached (D), sled components (E), laptop (F), panel meter (G), power supplies (H), rotary switch (I), terminal blocks (J), A/D converter (K), cross supports (L), power strip (M), battery (N) powering the foot trigger (O), and weights (P). The laptop and weights are not pictured on the left.
50
Figure 33: Close-up schematic of the finalized pendulum device. The locations of the rotary potentiometer (A), pendulum arm (B), original pendulum drop mass (C), additional pendulum drop mass attached (D), linear potentiometer (E), linear bearings (F), sled (G), adjustable mounting mechanism (H), coil spring (I), sled lever (J), and cam follower (K) are as noted.
51
3.3 Data Collection Technique
In order to collect the voltage data from the potentiometers, LabVIEW
SignalExpress (National Instruments Corp., Austin, TX) was used. A referenced single
ended (RSE) voltage setup with two analog voltage inputs (rotary potentiometer and
linear potentiometer) was implemented with a digital trigger (foot trigger). Two seconds
of potentiometer data were collected (30,000 samples at 15,000 hertz) for a single
impaction. The average duration between the release of the pendulum drop mass and the
end of an impaction was 0.7 seconds, so two seconds was sufficient time to gather the
entire impaction history information. The sampling rate was chosen for enough samples
to be collected to filter the signal and eliminate any unwanted noise. Once collected in
LabVIEW, the raw voltage data were exported to Excel for permanent record.
With the raw voltage data saved in Excel, a custom MATLAB script was written
to automate the process of converting the voltage information into the desired energy
information. The script could analyze a single file or a folder of Excel files. To start, both
potentiometer signals were filtered through an 11-point moving average filter to attenuate
the high frequency noise components. The appropriate number of points to average was
chosen by analyzing the preliminary impaction information. An 11-point moving average
filter minimized the amount of memory required for computation, while effectively
removing the high frequency noise signals. After filtering, the voltage history from the
rotary potentiometer was converted into linear position history. The linear position
history was plotted for graphical representation of the impaction, so that the user could
identify any possible signs of error. The maximum slope (derivative) of the linear
position prior to impact was found, and was used to calculate the pre-impact kinetic
energy of the pendulum drop mass. Next, the linear potentiometer voltage information
was converted into linear displacement information. With the known spring constant, the
maximum linear displacement was used to calculate the post-impact sled energy. The
difference between the calculated pre-impact kinetic energy and post-impact sled energy
52
was the energy absorbed by the specimen during impaction. The specimen number, pre-
impact kinetic energy, post-impact sled energy, and energy absorption information were
also automatically saved in an Excel format for permanent record and future user
manipulation. Again, the entire process is automated using the custom MATLAB script,
so meaningful results could be seen and saved in just a few seconds. With the final
pendulum device and data collection technique established, the device and methodology
were then ready for validation.
53
CHAPTER 4: VALIDATION STUDIES
This chapter discusses the studies that were conducted to validate the final version
of the pendulum device and its associated methodology. Two major validation studies
were conducted. The first validation study used foam specimens, which allowed for a
more controllable impaction and additional measurement ability. The second validation
study used foam and cadaver leg specimens, which was conducted to bridge between the
foam specimen validation study and contemplated in vivo IAF experiments. The
validation studies and corresponding results are outlined in the following sections.
4.1 Foam Specimen Validation Study
The foam specimen validation study used polyurethane foam (density of 240
kilograms per cubic meter) surrogates as stand-ins for the porcine hock specimens. To
maintain a more controlled impaction, an aluminum specimen indenter was machined
using a HAAS Precision Collet Lathe, to temporarily replace the tripod anchorage
system. The specimen indenter would indent the foam during impaction, rather than
causing fracture. This provided the ability to visualize and quantify the damage received
by the specimen. The experimental methods and results of the foam specimen validation
study were as follows.
4.1.1 Experimental Methods
Nine foam specimens were cut to size (2.5 x 3.8 x 3.8 centimeters) to
conveniently fit on the specimen sled for impaction. The orientation of the indenter and
foam specimens are shown in Figure 34. Each foam specimen could be impacted twice
on the same face—once on the bottom and once on the top. Foam specimens were
impacted six times at each of the three predetermined drop angles (20°, 40°, 60°). The
previously described data collection technique was utilized to gather the pre-impact
kinetic energy and post-impact sled energy information. As before, the energy absorption
54
was defined as the difference between the pre- and post-impact energies. After impaction,
the foam’s penetration depth caused by the indenter was measured with a digital caliper.
Figure 34: Pendulum device with indenter rigidly fixed with the mounting mechanism and foam specimen sitting on the low-friction sled.
Nine specimens of the same dimensions were also cut for indentation using an
MTS 858 Bionix materials testing machine (MTS Systems Corp., Eden Prairie, MN).
Again, each specimen was indented twice with the same indenter. The penetration depths
measured from the first nine specimens were mimicked using the displacement control of
the MTS, so the penetration depths between the first and second nine specimens would be
similar. After indentation, penetration depths were again measured with a digital caliper.
The force-displacement information was collected during the MTS indentations. The area
under the force-displacement curve was calculated to find the energy absorbed by the
55
specimen. The energy absorption calculations between the potentiometer data and the
MTS data were compared to investigate correlation.
4.1.2 Experiment Results
Both qualitative and quantitative experimental results from the foam specimen
validation study are presented and discussed in this section. The qualitative penetration
depth results of the foam specimens post-impact are illustrated in Figure 35. Penetration
depth can be thought of as a more conceptual representation of energy absorbed by a
foam specimen. It is logical to say that an increase in energy absorption would cause an
increase in penetration depth. From basic visual observation, it is evident that a larger
drop angle resulted in greater penetration depth, as was expected. The quantitative results
(Figure 36), from the potentiometer and MTS data, show the extremely high correlation
between the energy absorption calculations. The energy absorption calculation from the
potentiometer data incorporates the energy absorbed by the entire system (i.e.,
indentation, friction, and the pendulum device itself). The energy absorption calculation
from the MTS data only included the energy due to indentation of the specimen. Because
of this difference, the energy absorption calculation from the potentiometer data was
slightly greater than that from the MTS data. These results show that the technique
surrounding the energy absorption calculation is indeed valid and accurately represents
the amount of energy absorbed by a specimen during an impaction. However, does the
foam specimen validation study really apply to the case of cadaver leg or in vivo porcine
IAFs? To draw a connection between the foam specimen impactions and porcine IAF
impactions, another validation study was carried out.
56
Figure 35: Post-impact foam specimens with visual indentations. Each specimen was impacted twice—once on the bottom and once on the top. The top six impactions were conducted at an initial drop angle of 20°, the middle six at 40°, and the bottom six at 60°. For some impactions at the low drop angle, two indentations were evident in a single impaction. At the higher drop angles, two indentations did not occur. The penetration depth clearly increased with an increase in drop angle.
Figure 36: Relationship between energy absorption calculations from the potentiometer data and MTS data.
57
4.2 Motion Capture Validation Study
The motion capture validation study used two types of specimens: (i)
polyurethane surrogates like those in the previous validation study, and (ii) cadaver leg
porcine hock specimens. A motion capture system (Oqus Camera Series, Qualisys AB,
Gothenburg, Sweden) was utilized to provide optical kinematic measurements and high-
speed digital video of specimen impactions. This information was gathered to link the
validity of the foam specimen impactions to those for cadaver leg (and in vivo)
impactions. The experimental methods and results of the motion capture validation study
were as follows.
4.2.1 Experimental Methods
Similar to the previous validation study, nine foam specimens were cut to size
(2.5 x 3.8 x 3.8 centimeters) for impaction. Three foam specimens were impacted at each
of the three predetermined drop angles (20°, 40°, 60°). In addition, the same methodology
was used to gather the pre-impact kinetic energy, post-impact sled energy, energy
absorption, and penetration depth measurements. Four motion capture cameras were
positioned to track information from markers that were placed on the pendulum drop
mass, foam specimen, and tripod anchorage system. The marker on the tripod anchorage
system was equivalent to having one on the mounting mechanism, since they are rigidly
fixed to each other. The advantage to having the marker on the tripod anchorage system
instead of on the mounting mechanism was related to camera window size. With a
smaller window size (i.e., a smaller field of view), more data could be captured at a
higher sampling rate, which was desirable given the rapid nature of the impactions. The
marker on the foam specimen was only used for visual assurance of the specimen’s
position, since the foam specimen was not rigidly fixed to the pendulum. A hand trigger
was used to begin data collection, similar to the foot trigger for the pendulum device.
Once the motion capture system was properly positioned and ready, foam specimen
58
impactions were conducted. Two seconds of potentiometer data, 3D position data, and
high-speed digital video were all simultaneously collected via the previously discussed
data collection technique and the motion capture system. A sampling rate of 4 kilohertz
was chosen for the motion capture system, because that was the maximum possible rate
for the selected window size.
In succession to the foam specimen impactions, five cadaver leg porcine hock
specimens were impacted at a drop angle of 67°. This drop angle corresponded to an
initial gravitational potential energy measurement of 35 joules, which was representative
of preliminary IAF impactions in vivo. The motion capture device remained in position
for the cadaver leg impactions. However, since a marker could not meaningfully be used
if attached to the specimen, markers were only placed on the pendulum drop mass and
tripod anchorage system. As in the foam specimen impactions, two seconds of
potentiometer data, 3D position data, and high-speed digital video were all
simultaneously collected for the cadaver leg impactions.
For both types of specimens, the potentiometer data were used to calculate the
respective energy values, as was done in the foam specimen validation study. The 3D
position data gave the position history of the pendulum drop mass and mounting
mechanism. Once the 3D position data were exported to Excel, basic manipulation of the
data was conducted to extract the desired information. From the pendulum drop mass
position data, the linear velocity could be found to calculate the time history of the mass’s
kinetic energy. From the mounting mechanism position data and known spring constant,
a time history of the sled energy could be calculated. From these two energy
measurements, the energy absorption value could then be found. The energy
measurements from the 3D position data were compared to the measurements from the
potentiometer data. The initial length or distance between the pendulum drop mass and
mounting mechanism markers was also found and defined as Li. During the course of a
foam impaction, the initial length would decrease as the indenter was penetrating the
59
specimen. The final length between the two markers was found and defined as Lf. The
metric for the difference between initial and final lengths, !L, was designated as the
specimen deformation. Figure 37 illustrates the configuration of the four motion capture
cameras, the associated markers, and the specimen deformation metric. In the foam
impaction cases, it was assumed that the specimen deformation metric would closely
resemble the penetration depth measurement. The relationship between these two was
analyzed to confirm the aforementioned assumption. If the specimen deformation metric
were equivalent to penetration depth in the foam impactions, then it could also be applied
to the cadaver leg impactions for comparison. The high-speed digital video was also used
to visually confirm the 3D position data and to provide an understanding of the
specimen’s motion history throughout impaction.
Figure 37: Motion capture system setup for the foam surrogate specimens (top) and cadaver leg porcine hock specimens (bottom). The markers and specimen deformation metric are labeled.
60
4.2.2 Experiment Results
Comparison between the energy calculations from the 3D position data and
potentiometer data can be seen in Table 2 for the foam impactions. It is evident that the
energy calculations from the motion capture information are similar to the energy
calculations from the potentiometer information. This confirms that the pendulum device
instrumentation and data collection technique are accurate.
Table 2: Energy measurement results from the potentiometer and motion capture data for each foam specimen, where KE is the kinetic energy, SE is sled energy, and EA is energy absorption.
Visual observations of the foam specimens post-impact were also consistent with
the qualitative penetration depth results seen in the foam specimen validation study. The
comparison between the penetration depth and specimen deformation measurements for
each of the three drop angles can be seen in Table 3. It is evident that the specimen
deformation and penetration depth measurements are indeed approximately equivalent.
Therefore, the specimen deformation measurement meaningfully applied to the cadaver
leg impactions.
61
Table 3: Foam specimen penetration depth and specimen deformation measurements.
From the motion capture information, the energy and specimen deformation
histories for a representative foam impaction are shown in Figure 38. It can be seen that
the majority of the pre-impact kinetic energy was transferred to the specimen and sled
early during the impaction event. The kinetic energy was not entirely transferred to the
specimen and sled until the sled’s forward motion was halted and reached its maximum
displacement. The sled energy gradually increased from the initial time of impaction until
the sled’s forward motion was halted upon reaching its maximum displacement. The
specimen deformation history revealed the interesting kinematic behavior of the foam
specimen during an impaction. The majority of the specimen deformation occurred
during the beginning of impaction. Near the instant of the sled’s maximum displacement,
the maximum specimen deformation was also reached. The double-peak feature of the
specimen deformation curve was explained from the high-speed digital video. The foam
specimen was being penetrated during the initial peak, while the low-friction sled was
still beginning to accelerate forward. After the first peak, the sled began to move slightly
62
faster than the pendulum drop mass. Due to the measurement technique of the specimen
deformation metric, a decrease in specimen deformation was seen on the history curve.
With the coil spring constantly resisting the sled’s forward motion, the pendulum drop
mass started to reconnect with the foam specimen. This was the cause of the second peak,
where the maximum specimen deformation occurred. Shortly there after, the sled’s
forward motion was halted by the coil spring. The high-speed digital video confirmed the
characteristics of the energy and specimen deformation histories in the foam specimen
impactions. Again, it is important to note that the majority of the penetration depth
occurred during the beginning of impaction, which the first specimen deformation peak
alluded to.
Figure 38: Kinetic energy, sled energy, and specimen deformation histories for a representative foam impaction.
63
For the cadaver leg specimen impactions, the energy and specimen deformation
histories of a representative impaction are shown in Figure 39. The cadaver leg impaction
results were strikingly similar to the foam impaction results. The majority of the pre-
impact kinetic energy was again transferred to the specimen and sled during the
beginning of impaction. The sled energy again gradually increased from the initial time
of impaction until the sled’s forward motion was halted. The majority of the specimen
deformation again occurred during the beginning of impaction. However, the double-
peak was not as dramatic for the foam impaction. Following the first specimen
deformation peak, the deformation curve gradually increased until the sled’s forward
motion was halted by the coil spring. The high-speed digital video also confirmed the
characteristics of the energy and specimen deformation histories in the cadaver leg
specimen impactions. From the video information, it also appeared that the initial IAF of
the cadaver leg specimen was created during the beginning of impaction.
The motion capture results from the foam and cadaver leg impactions revealed
many similarities. The energy history information was nearly identical, and the specimen
deformation histories were also very comparable. In addition, both the foam and cadaver
leg specimens experienced the majority of their damage during the initial portions of
impaction. Based on the similarities seen in the motion capture validation study, the
energy absorption measurement technique that was validated in the foam specimen
validation study could indeed be reasonably applied to cadaver leg and in vivo
impactions. With the energy absorption measurement technique successfully validated, in
vivo impactions could finally be conducted. The materials and methods for the in vivo
fracture specimens are set forth in the next chapter.
64
Figure 39: Kinetic energy, sled energy, and specimen deformation histories for a representative cadaver leg porcine hock impaction.
65
CHAPTER 5: MATERIALS AND METHODS FOR IN
VIVO FRACTURE ANIMALS
This chapter outlines the materials and methods that were used for in vivo porcine
IAF creation, work still ongoing at the time of this writing. The steps taken in preparing
for the live animal impaction are discussed, as well as the surgical procedure for
attaching the tripod anchorage system, the impaction technique for IAF creation, and the
post-impact surgical and animal care procedures.
Eleven Yucatan miniature pigs (IACUC-approved, ACURF #1007141) have been
impacted to date using the pendulum device and data collection technique. Animals #1-4
were impacted in the spring of 2011. Animals #5-11 were impacted in the spring of 2012.
For the remainder of the materials and methods, focus will be placed on those animals
that have already been impacted. Future plans with the ongoing in vivo study will be
discussed in Chapter 6.
5.1 Preparation for the Live Animal Impaction
The first four Yucatan miniature pigs (age of approximately two years and weight
of 55 kilograms) were obtained from Sinclair Bio Resources (Auxvasse, MO). The next
seven Yucatan miniature pigs (age of approximately two years and weight of 90
kilograms) were obtained from Exemplar Genetics (Sioux Center, IA). The animals were
maintained according to University of Iowa animal care guidelines. Experimental
surgeries were performed by an experienced orthopaedic surgeon, while using inhalation
anesthesia and aseptic technique.
Before the surgical procedure commenced, appropriate steps were conducted to
ensure the pendulum device and data collection hardware were operating correctly. The
voltages from the power supplies were checked to confirm that they were supplying each
of the potentiometers with 10 volts. With the correct voltage being supplied, the voltages
from the resting positions of each potentiometer were recorded. These voltages were
66
checked with previous records, as they were also needed for the energy calculations. The
sled lever was used to test the linear potentiometer by moving the low-friction sled.
Similarly, the pendulum arm was rotated to test the rotary potentiometer. The foot trigger
was also tested multiple times to ensure the batteries were still supplying 3 volts when
triggered.
5.2 Pre-Impact Surgical Procedure
After the animal was prepared and the hardware tested, the orthopaedic surgeon
performed surgery to attach the tripod anchorage system to the hock joint. The surgical
procedure was similar to that discussed above for the tripod anchorage system section—
attention is directed to Figure 11 for reference. The three external fixator pins (6
millimeter diameter, Orthofix Inc., Lewisville, Texas) were used to fix the aluminum
impact interface to the talus. Two external fixator pins were directly inserted into the
talus, inferomedially. The third external fixator pin was inserted through the talocalcaneal
joint into the talus. The inserted external fixator pins effectively created a tripod
configuration, hence the nomenclature of the anchorage system. Next, the leg holder
device was attached to the tibia. The distal tibia plate was attached using two plugs and a
cortical screw, which were inserted into the anterior distal tibial surface. The plate’s ball-
and-socket joint allowed the leg holder shaft to be properly oriented for impaction. The
leg holder shaft gripped another external fixator pin that was inserted through the
proximal tibia. This attachment had the dual purpose of maintaining the inclination of the
tibial shaft (approximately 45°) and maintaining the orientation of the leg holder shaft in
the direction of impact mass travel. After the tripod anchorage system was successfully
attached to the joint, the stress-rising saw cut was created using an oscillating bone saw
(Stryker Corp., Kalamazoo, MI) to ensure a consistent fracture pattern.
67
5.3 Intraarticular Fracture Creation
Once the tripod anchorage system was attached properly to the animal, the
pendulum device and data collection technique were used to create the IAF insult. The
surgical utensil tables and fluoroscopy equipment were removed from the immediate
surgical field, while the pendulum device and data collection hardware were introduced.
The specimen table was placed over the operating table and under the specimen joint.
The leg holder shaft of the tripod anchorage system was positioned into the mounting
mechanism and adjusted until the impact interface was centered about the pendulum drop
mass. Once the tripod anchorage system and animal joint were properly positioned, the
mounting mechanism bolts were tightened to lock the leg holder shaft in place. Figure 40
shows the pendulum device in the surgical environment with the animal leg properly
mounted prior to fracture. The data collection software was commanded to collect two
seconds of potentiometer voltage data when triggered. The pendulum drop mass was
raised to the desired drop angle by observing the panel meter information. The foot
trigger was pressed, and the pendulum drop mass was released from the correct angle.
The pendulum drop mass was then manually caught following impaction, so as to prevent
a second impaction. After impaction, the orthopaedic surgeon observed the damage to the
animal (Figure 41). If it was determined that the initial IAF attempt was not successful,
another impaction was performed. Often, in the case of another impaction, the drop angle
was increased to improve the chances for IAF occurrence. The potentiometer data from
each impaction were recorded in LabVIEW and exported to Excel. To reiterate, the
voltage output from the rotary potentiometer corresponded to the angular position of the
pendulum arm. From the angular position information, the linear velocity could be found
and used to calculate the kinetic energy of the pendulum drop mass immediately before
impact (pre-impact kinetic energy). The voltage output from the linear potentiometer
corresponded to the linear position of the low-friction sled. From the linear position
information and the known spring constant, the energy required for the coil spring to stop
68
the low-friction sled’s forward progress (post-impact sled energy) could be calculated.
The energy absorption measurement was then the difference between the pre-impact
kinetic energy and post-impact sled energy. As previously discussed, a custom MATLAB
script automatically analyzed the data and calculated the kinetic energy, sled energy, and
energy absorption measurements. Again, the final information was exported to Excel for
permanent record. Following a successful fracture insult, the pendulum device and data
collection hardware were removed from the immediate surgical field, and then the
surgical utensil tables and fluoroscopy equipment were reintroduced.
Figure 40: Pendulum device in vivo, including a rotary potentiometer that measured the pendulum’s linear velocity (A), the linear potentiometer that measured the sled’s linear displacement (B1), the specimen anchorage system (B2), the low-friction sled on linear bearings (B3), and the coil spring that resisted the sled’s forward motion (B4).
69
Figure 41: Orthopaedic surgeon checking for post-impact fracture.
5.4 Post-Impact Surgical Procedure and Animal Care
Immediately following the fracture insult, the orthopaedic surgeon prepared to
anatomically reduce the fracture fragments. A 2.1-millimeter thick tibial plateau leveling
osteotomy (TPLO) plate, specifically designed for veterinary surgery (Synthes Vet, West
Chester, PA), was surgically fixed across the distal tibial fracture with six 2.7-millimeter
bone screws. Once the TPLO plate was in place (Figure 42), the open incisions were
sutured closed. Following suture, the animal joint was put in a surgical cast for a week,
and the animals were again maintained according to University of Iowa animal care
guidelines. In addition to the radiographs taken before fracture insult (baseline) and
immediately after fracture, radiographs were also taken after anatomical reduction and at
various time points post-fracture (1, 2, 4, 8, and 12 weeks). Mainly, the radiographs
allowed the orthopaedic surgeon to identify the characteristics of each fracture and
observe the corresponding bone healing progression.
70
Figure 42: Tibial plateau leveling osteotomy plate across the post-impact IAF insult.
71
CHAPTER 6: RESULTS & DISCUSSION
This chapter presents the results from the eleven animals studied to date,
including IAF creation and energy measurement results. The animals were divided into
two groups based on the date of impactions and the associated experimental variables.
6.1 Live Animals #1-4
6.1.1 Intraarticular Fracture Creation Results & Discussion
Being the first in vivo impaction that was conducted, Animal #1 not unexpectedly
experienced some complications. The first attempt at fracturing the joint was not
successful using a drop angle that corresponded to 20 joules of initial gravitational
potential energy. Two additional attempts at a 30-joule impact and a fourth attempt at 40
joules were needed to finally produce fracture. The need for multiple impactions could
possibly be explained due to inexperience. Mainly, the necessary drop angle to produce a
sufficient magnitude of pre-impact energy to cause fracture was not known. From the
four fracture attempts in Animal #1, clearly an initial gravitational potential energy
greater than 30 joules would be required in the porcine animals to follow. To account for
a larger energy magnitude, the coil spring (spring #2 discussed in Chapter 2) was
upgraded to a spring that possessed a larger energy capacity (spring #3). The pendulum
drop mass that was used for Animal #1 allowed for a maximum initial gravitational
potential energy of only 38 joules at 90° (maximum angle for consistent manual release).
Any needed increase in pre-impact energy would not be possible, so the pendulum drop
mass was increased. It should also be noted that the severity of fracture following the
fourth fracture attempt was greater than what was preferred. This could have been related
to the multiple impactions that were delivered to the animal joint. The one-day post-
surgical reduction radiograph showing the fracture and TPLO plate from Animal #1 can
be seen in Figure 43.
72
Figure 43: One-day post-fracture radiograph of the fracture insult and TPLO plate from Animal #1.
After the pendulum system upgrades were completed, Animals #2-4 were
impacted. Animals #2 and #3 did not experience the same complications as Animal #1.
Both animals successfully fractured with a single impaction with an initial gravitational
potential energy of 33 joules. This was much more encouraging than Animal #1. From
these results, it appeared that the additional mass and spring #3 were sufficient for
fracture creation. When Animal #4 was impacted under the same conditions as Animals
#2 and #3, however, fracture did not occur after a first impaction with an initial
gravitational potential energy of 33 joules. After a second impaction at 40 joules,
however, fracture was successfully created. The post-surgical reduction radiographs from
Animals #2-4 were similar to the radiograph from the first animal. The severities of the
IAFs created in Animals #2-4 were less than the IAF created in Animal #1. Moreover, the
IAFs created in Animals #2-4 were morpholically consistent with human ankle fractures,
which was a central objective for the in vivo study.
73
6.1.2 Energy Measurement Results & Discussion
The experience with data collection and the energy measurement results for the
first four animals are discussed in this section. The associated variables and energy
measurement results for these animals are shown in Table 4.
Table 4: Energy measurement results for Animals #1-4.
Note: The associated variables are given, such as pendulum drop mass, spring number, whether or not the stopping mechanism was used, and the theoretical gravitational potential energy of impaction.
For Animal #1, the energy absorption measurement was 32 joules during the final
impact, for which fracture occurred. It was apparent from the spring energy measurement
that the maximum energy capacity of spring #2 was being reached. Therefore, the energy
absorption that was calculated could have been greater than the energy actually absorbed
physically. When comparing the kinetic energy value to the theoretical potential energy
value, the kinetic energy was greater, which was not expected. This could be explained
by analyzing the voltage outputs from the potentiometers (Figure 44). The initial reading
of the rotary potentiometer indicated that the sliding contact reached its zero voltage
position at such a high drop angle (>100°). Therefore, the theoretical potential energy
value was not as high as it should have been, causing it to be less than the kinetic energy
value. These energy measurement results confirmed the initial need for upgrades based
on the IAF creation results. The additional mass would eliminate the need for excessive
74
drop angles, thus allowing for more accurate energy measurements and for a more user-
friendly release of the pendulum drop mass. The upgraded spring would also allow for
the needed increase in energy capacity.
Figure 44: Linear and rotary potentiometer voltage outputs for Animal #1. The linear displacement graph shows the sled beginning to translate at around 0.9 seconds (when impaction initiated). The displacement remained at 4.4 centimeters because the stopping mechanism was used.
To reiterate, Animals #2-4 were impacted after the additional mass and spring #3
were attached. Animals #2 and #3 had energy absorption measurements of 12 and 14
joules, respectively. The voltage outputs from the potentiometers did not experience the
75
same problem as in Animal #1 (Figure 45). For Animal #4, the energy absorption
measurement was 17 joules during the impact that induced fracture. It appeared that the
additional mass and updated spring were sufficient for the in vivo impactions, but other
improvements needed to be made before proceeding to fracture more animals in
subsequent in vivo impactions.
Figure 45: Linear and rotary potentiometer voltage outputs for Animal #2. The linear displacement graph shows the sled beginning to translate at around 0.8 seconds (when impaction initiated). The displacement remained at 3.3 centimeters because the stopping mechanism was used. These voltage outputs were similar to those from Animals #3 and #4.
76
6.1.3 Further Modifications
After observing the results and surgical procedure, further modifications were
made to facilitate more fluid system operation. To start, the stopping mechanism
experienced numerous problems, as discussed in Chapter 2. Consequently, it was decided
to remove the stopping mechanism to eliminate those complications. The data collection
hardware was upgraded and fixed to the pendulum, eliminating the need for extra cords
and the data collection cart (see Chapter 2 for details). The adjustable mounting
mechanism was also modified to allow for increased adjustability in the vertical and
transverse directions (see Chapter 2 for details). These modifications made the pendulum
device easier to operate, and allowed for increased animal maneuverability for the
orthopaedic surgeon. Once the modifications were established, future in vivo impactions
were conducted for an additional seven animals.
6.2 Live Animals #5-11
6.2.1 Intraarticular Fracture Creation Results & Discussion
Animals #5-11 were impacted after the modifications were implemented. Animals
#5 and #6 required single impactions with an initial gravitational potential energy of 45
and 46 joules, respectively, to create a successful fracture. Fracture for Animal #7
required two impactions at 41 joules of initial gravitational potential energy. Animals #8
and #9 only required single impactions at 46 and 41 joules, respectively. For Animal #9,
the initial gravitational potential energy was intended to be 45 joules. However, the drop
angle of the pendulum arm was incorrect, due to user error in the release. Animal #10
needed to be impacted twice, at 45 and 46 joules, to achieve fracture. Following the first
impaction, the orthopaedic surgeon re-adjusted the tripod anchorage system, because it
was not properly aligned in the pendulum device. This was believed to be the cause of the
first failed impaction. Therefore, the energy delivery magnitude was not deliberately
increased for the second impaction. Animal #11 exhibited a successful fracture following
77
a single impaction of 46 joules. The post-surgical reduction radiographs for Animals #5-
11 were similar to those of Animals #1-4. Similar to Animals #2-4, the IAFs created in
Animals #5-11 were morphologically similar to human ankle fractures.
6.2.2 Energy Measurement Results & Discussion
The data collection technique and energy measurement results for Animals #5-11
are discussed in this section. The associated variables and energy measurement results for
Animals #5-11 can be seen in Table 5.
Table 5: Energy measurement results for Animals #5-11.
Note: The associated variables are given, such as pendulum drop mass, spring number, whether or not the stopping mechanism was used, and the theoretical gravitational potential energy of impaction.
For Animal #5, the energy absorption measurement was 28 joules during the final
impact that induced fracture. When observing the spring energy measurement, the
maximum energy capacity was being approached. The energy absorption measurement
increased more than the input kinetic energy when compared to the results from Animal
#4. The spring energy capacity being approached could have been the explanation for the
higher energy absorption measurement. Therefore, it was decided to make a final spring
78
upgrade (spring #4) that would eliminate any possibility of reaching the maximum energy
capacity.
After the final spring upgrade, Animals #6-11 were impacted. Typical voltage
outputs from the potentiometers can be seen in Figure 46.
Figure 46: Linear and rotary potentiometer voltage outputs for Animal #6. The linear displacement graph shows the sled beginning to translate at around 0.9 seconds (when impaction initiated). The stopping mechanism was not used, so the linear displacement returned to zero. These voltage outputs were similar to those from Animals #7-11.
79
Animal #6 had an energy absorption measurement of 21 joules. For Animal #7,
the energy absorption measurement was 20 joules during the final impact that induced
fracture. Animals #8 and #9 had energy absorption measurements of 21 and 20 joules,
respectively. After the final impact, Animal #10 had an energy absorption measurement
of 23 joules. For Animal #11, the energy absorption measurement was 20 joules.
6.3 Limitations and Potential Solutions
Despite the success of the pendulum device and data collection technique, there
are some limitations that need to be addressed. First, the energy absorption measurement
does not exclusively measure the fracture energy. Since the potentiometer data only
provides the pre- and post-impact energies, the energy absorption measurement also
includes energy that dissipates throughout the impaction to various avenues, including
soft tissue deformation, friction, and the pendulum device itself. Further investigation of
the motion capture data, should provide more information about the energy dissipation in
real-time. This would allow for a better understanding of energy that solely corresponds
to the fracture energy.
The user error in the release of the pendulum drop mass is another limitation of
the pendulum device. A potential solution that would eliminate this source of error is a
release mechanism. The release mechanism would ensure that the drop height of the
pendulum drop mass was consistent between animals.
80
CHAPTER 7: CONCLUSION
This chapter discusses conclusions that can be drawn from the IAF creation and
energy absorption measurement results of the first eleven animals that have been
impacted to date in the in vivo porcine IAF study. In addition, future (funded) research
that intends to use the described pendulum device and data collection technique will also
be discussed. Finally, implications of this work are discussed applicable to the broader
scope of other impaction devices that might be utilized in the future.
7.1 Intraarticular Fracture Creation
Looking at the success rate of IAF creation in the eleven animals impacted to
date, fracture was successfully achieved in every instance. However, it was desirable for
only a single impaction to be required to create fracture, which occurred in seven out of
the eleven animals (64%). This success rate is low, but most of the failure instances were
due to the learning process of developing such a novel IAF model. Animal #1 and #4
required multiple impactions because the magnitude of the initial gravitational potential
energy needed to create fracture was not known. For the final six animals, the variables
remained constant, except for the initial gravitational potential energy of Animal #7.
Animal #7 was the only animal to intentionally receive an initial gravitational potential
energy less than 45 joules, which resulted in the need for another impaction to create
fracture. Animal #10 was the other animal that required a second impaction. Again, this
apparently was due to surgical error in positioning the animal, which seemingly could be
avoided in future impactions. If Animal #10 is disregarded due to this surgical error, the
success rate for the other five animals was 80%. With additional practice and experience,
minor inaccuracies like the surgical error in Animal #10 could potentially be avoided,
leading the success rate of the IAF model to further improve. The current pendulum
device and its modifications proved to be effective for creating fracture. Therefore, the
desirable variables that would likely result in consistent fracture after a single impaction
81
in the future are: 5.82-kilogram drop mass, fourth spring, no stopping mechanism, and a
drop angle corresponding to 45 joules of initial gravitational potential energy.
7.2 Energy Absorption Measurement
The energy absorption measurement results for the fractures in the first eleven
animals exhibited some variability. The range of energy absorption measurements during
fracture of the eleven animals was 11.7-31.8 joules, with a mean and standard deviation
of 20.8 ± 5.7 joules. Both the range and the variance are quite large. However, if the
animals are stratified based on the impaction variables, the sources of variability can be
revealed. During fracture, the first five animals had an energy absorption measurement
ranging from 11.7-31.8 joules, with a mean and standard deviation of 20.6 ± 8.9 joules.
For the fractures in Animals #6-11, the energy absorption measurements ranged from
19.7-22.6 joules, with a mean and standard deviation of 20.9 ± 1.1 joules. From this
information, it is evident that the large range and variability among the eleven animals
resulted from experience in the first five animals. The variability in the first five animals
could be explained by the variables that were being adjusted between animals to improve
the consistency of fracture creation. After appropriate values of these variables were
identified, Animals #6-11 demonstrated gratifyingly consistent energy absorption
behavior. The consistency seen in Animals #6-11 further indicates that the data collection
technique and energy absorption measurement are appropriate for the in vivo IAF model.
In addition to the consistency of Animals #6-11, the magnitude of energy
absorption measurements was also in the range that is expected. On average, the energy
absorption measurements were approximately 52% of the pre-impact kinetic energy
values. This agreed with the values seen in the motion capture validation study, which
had energy absorption measurements that were approximately 50 and 51% of the pre-
impact kinetic energy values for the foam and cadaver leg porcine hock animals,
respectively. The magnitude of energy absorption measurements for Animals #6-11 also
82
agreed with the work from Abdel-Wahab et al. [49], which showed that the energy
absorption of cortical bone fracture in a controlled Izod impact test was approximately
51%. Based on these results, the magnitude of energy absorption measurements seems to
be quite accurate for in vivo IAF work.
7.3 Future (Funded) Research
The ongoing in vivo porcine IAF study plans to impact another six animals in the
next month. The current pendulum device and data collection technique will be used to
collect the energy absorption measurements for these animals, as was done previously.
Funding sources have committed support for a total of 48 animals to be impacted by the
end of next year as part of the ongoing in vivo porcine IAF study. Once that study has
been completed, an effective large animal IAF model will be established. The progression
of chondrocyte damage associated with the established IAF model will allow for the
testing of possible biological adjunct therapies for posttraumatic OA treatment. Once
effective posttraumatic OA treatments are explored, they could possibly also be applied
to treat other manifestations of OA.
7.4 Future of the Energy Absorption Technique
The current pendulum device and data collection technique are not limited to IAF
insult in porcine animals. The described methodology could easily be applied to other
large animals. Furthermore, its use could be applied to other impaction experiments, such
as blunt impaction experiments, with a few minor adjustments to the physical pendulum
device.
The idea of performing energy absorption measurement should also be considered
and applied to other blunt impaction and fracture insult devices. As discussed in Chapter
1, most of the published blunt impaction and fracture insult devices can only calculate the
initial gravitational potential energy of impaction. If future blunt impaction and fracture
insult devices could incorporate energy absorption measurements, the actual amount of
83
energy damaging the animal could be controlled for. This would provide novel
information associated with the pathomechanics of the induced injury.
In conclusion, the described pendulum device and data collection technique have
been validated and effectively utilized in an ongoing in vivo porcine IAF study. The IAF
creation results showed that the device was successful in creating fractures, while the
energy absorption measurement results showed that the data collection technique was
successful in quantifying the fracture insult. The methodology of this work is being
applied to future animals in the ongoing in vivo porcine IAF research and seemingly can
be implemented in future blunt impaction and fracture insult devices.
84
APPENDIX
Table A1: Dates and types of impaction tests that were conducted with the pendulum device.
Note: The associated variables are given, such as linear potentiometer number, rotary potentiometer number, pendulum drop mass, spring number, indenter or mounting mechanism number, and whether or not the stopping mechanism was used. The last column describes the purpose of each impaction test that was performed.
85
Figure A1: Circuit diagram of the data collection instrumentation and wiring. The letters correspond to the color of wire. The rotary switch has four paired nodes (nodes vertically opposite each other) that correspond to a position that could be cycled through and seen on the panel meter.
86
REFERENCES
1. Osteoarthritis: National clinical guideline for care and management in adults 2008: London.
2. Buckwalter, J.A. and H.J. Mankin, Articular cartilage: degeneration and
osteoarthritis, repair, regeneration, and transplantation. Instr Course Lect, 1998. 47: p. 487-504.
3. Buckwalter, J.A. and J. Martin, Degenerative joint disease. Clin Symp, 1995.
47(2): p. 1-32. 4. Buckwalter, J.A., J. Martin, and H.J. Mankin, Synovial joint degeneration and the
syndrome of osteoarthritis. Instr Course Lect, 2000. 49: p. 481-9. 5. Buckwalter, J.A. and J.A. Martin, Osteoarthritis. Adv Drug Deliv Rev, 2006.
58(2): p. 150-67. 6. Ponseti, I.V., J.A. Maynard, S.L. Weinstein, E.G. Ippolito, and J.G. Pous, Legg-
Calve-Perthes disease. Histochemical and ultrastructural observations of the epiphyseal cartilage and physis. J Bone Joint Surg Am, 1983. 65(6): p. 797-807.
7. Brown, T.D., R.C. Johnston, C.L. Saltzman, J.L. Marsh, and J.A. Buckwalter,
Posttraumatic osteoarthritis: a first estimate of incidence, prevalence, and burden of disease. J Orthop Trauma, 2006. 20(10): p. 739-44.
8. Kellgren, J.H., Osteoarthrosis in patients and populations. Br Med J, 1961.
2(5243): p. 1-6. 9. Kirkley, A., T.B. Birmingham, R.B. Litchfield, J.R. Giffin, K.R. Willits, C.J.
Wong, B.G. Feagan, A. Donner, S.H. Griffin, L.M. D'Ascanio, J.E. Pope, and P.J. Fowler, A randomized trial of arthroscopic surgery for osteoarthritis of the knee. N Engl J Med, 2008. 359(11): p. 1097-107.
10. Lee, P., P.J. Rooney, R.D. Sturrock, A.C. Kennedy, and W.C. Dick, The etiology
and pathogenesis of osteoarthrosis: a review. Semin Arthritis Rheum, 1974. 3(3): p. 189-218.
11. Sharma, L., D. Kapoor, and S. Issa, Epidemiology of osteoarthritis: an update.
Curr Opin Rheumatol, 2006. 18(2): p. 147-56. 12. Boden, B.P., F.T. Sheehan, J.S. Torg, and T.E. Hewett, Noncontact anterior
cruciate ligament injuries: mechanisms and risk factors. J Am Acad Orthop Surg, 2010. 18(9): p. 520-7.
13. Roos, H., T. Adalberth, L. Dahlberg, and L.S. Lohmander, Osteoarthritis of the
knee after injury to the anterior cruciate ligament or meniscus: the influence of time and age. Osteoarthritis Cartilage, 1995. 3(4): p. 261-7.
14. Nikolic, D., Z. Jovanovic, Z. Popovic, R. Vulovic, and M. Mladenovic, Primary
surgical treatment of war injuries of major joints of the limbs. Injury, 1999. 30(2): p. 129-34.
87
15. Isaac, D.I., E.G. Meyer, and R.C. Haut, Chondrocyte damage and contact pressures following impact on the rabbit tibiofemoral joint. J Biomech Eng, 2008. 130(4): p. 041018.
16. Reginster, J.Y., The prevalence and burden of arthritis. Rheumatology (Oxford),
2002. 41 Supp 1: p. 3-6. 17. Leigh, J.P., W. Seavey, and B. Leistikow, Estimating the costs of job related
arthritis. J Rheumatol, 2001. 28(7): p. 1647-54. 18. Elders, M.J., The increasing impact of arthritis on public health. J Rheumatol
Suppl, 2000. 60: p. 6-8. 19. Gabriel, S.E., C.S. Crowson, and W.M. O'Fallon, Costs of osteoarthritis:
estimates from a geographically defined population. J Rheumatol Suppl, 1995. 43: p. 23-5.
20. Jackson, D.W., T.M. Simon, and H.M. Aberman, Symptomatic articular cartilage
degeneration: the impact in the new millennium. Clin Orthop Relat Res, 2001(391 Suppl): p. S14-25.
21. Lapsley, H.M., L.M. March, K.L. Tribe, M.J. Cross, and P.M. Brooks, Living with
osteoarthritis: patient expenditures, health status, and social impact. Arthritis Rheum, 2001. 45(3): p. 301-6.
22. Leardini, G., F. Salaffi, R. Caporali, B. Canesi, L. Rovati, and R. Montanelli,
Direct and indirect costs of osteoarthritis of the knee. Clin Exp Rheumatol, 2004. 22(6): p. 699-706.
23. MacLean, C.H., K. Knight, H. Paulus, R.H. Brook, and P.G. Shekelle, Costs
attributable to osteoarthritis. J Rheumatol, 1998. 25(11): p. 2213-8. 24. Murphy, L.B., C.G. Helmick, M.G. Cisternas, and E.H. Yelin, Estimating medical
costs attributable to osteoarthritis in the US population: comment on the article by Kotlarz et al. Arthritis Rheum, 2010. 62(8): p. 2566-7; author reply 2567-8.
25. Furman, B.D., J. Strand, W.C. Hembree, B.D. Ward, F. Guilak, and S.A. Olson,
Joint degeneration following closed intraarticular fracture in the mouse knee: a model of posttraumatic arthritis. J Orthop Res, 2007. 25(5): p. 578-92.
26. MacKenzie, E.J., M.J. Bosse, J.F. Kellam, A.R. Burgess, L.X. Webb, M.F.
Swiontkowski, R.W. Sanders, A.L. Jones, M.P. McAndrew, T.M. Patterson, and M.L. McCarthy, Characterization of patients with high-energy lower extremity trauma. J Orthop Trauma, 2000. 14(7): p. 455-66.
27. Segui-Gomez, M. and E.J. MacKenzie, Measuring the public health impact of
injuries. Epidemiol Rev, 2003. 25: p. 3-19. 28. McKinley, T.O., J. Borrelli, Jr., D.D. D'Lima, B.D. Furman, and P.V. Giannoudis,
Basic science of intra-articular fractures and posttraumatic osteoarthritis. J Orthop Trauma, 2010. 24(9): p. 567-70.
88
29. Hembree, W.C., B.D. Ward, B.D. Furman, R.D. Zura, L.A. Nichols, F. Guilak, and S.A. Olson, Viability and apoptosis of human chondrocytes in osteochondral fragments following joint trauma. J Bone Joint Surg Br, 2007. 89(10): p. 1388-95.
30. Kim, H.T., M.Y. Lo, and R. Pillarisetty, Chondrocyte apoptosis following
intraarticular fracture in humans. Osteoarthritis Cartilage, 2002. 10(9): p. 747-9. 31. Murray, M.M., D. Zurakowski, and M.S. Vrahas, The death of articular
chondrocytes after intra-articular fracture in humans. J Trauma, 2004. 56(1): p. 128-31.
32. Dang, A.C. and H.T. Kim, Chondrocyte apoptosis after simulated intraarticular
fracture: a comparison of histologic detection methods. Clin Orthop Relat Res, 2009. 467(7): p. 1877-84.
33. Muller, F.J., L.A. Setton, D.H. Manicourt, V.C. Mow, D.S. Howell, and J.C. Pita,
Centrifugal and biochemical comparison of proteoglycan aggregates from articular cartilage in experimental joint disuse and joint instability. J Orthop Res, 1994. 12(4): p. 498-508.
34. Setton, L.A., V.C. Mow, F.J. Muller, J.C. Pita, and D.S. Howell, Mechanical
behavior and biochemical composition of canine knee cartilage following periods of joint disuse and disuse with remobilization. Osteoarthritis Cartilage, 1997. 5(1): p. 1-16.
35. Ward, B.D., B.D. Furman, J.L. Huebner, V.B. Kraus, F. Guilak, and S.A. Olson,
Absence of posttraumatic arthritis following intraarticular fracture in the MRL/MpJ mouse. Arthritis Rheum, 2008. 58(3): p. 744-53.
36. Borrelli, J., Jr., M.E. Burns, W.M. Ricci, and M.J. Silva, A method for delivering
variable impact stresses to the articular cartilage of rabbit knees. J Orthop Trauma, 2002. 16(3): p. 182-8.
37. Borrelli, J., Jr., M.J. Silva, M.A. Zaegel, C. Franz, and L.J. Sandell, Single high-
energy impact load causes posttraumatic OA in young rabbits via a decrease in cellular metabolism. J Orthop Res, 2009. 27(3): p. 347-52.
38. Borrelli, J., Jr., K. Tinsley, W.M. Ricci, M. Burns, I.E. Karl, and R. Hotchkiss,
Induction of chondrocyte apoptosis following impact load. J Orthop Trauma, 2003. 17(9): p. 635-41.
39. Borrelli, J., Jr., M.A. Zaegel, M.D. Martinez, and M.J. Silva, Diminished
cartilage creep properties and increased trabecular bone density following a single, sub-fracture impact of the rabbit femoral condyle. J Orthop Res, 2010. 28(10): p. 1307-14.
40. Ewers, B.J., V.M. Jayaraman, R.F. Banglmaier, and R.C. Haut, Rate of blunt
impact loading affects changes in retropatellar cartilage and underlying bone in the rabbit patella. J Biomech, 2002. 35(6): p. 747-55.
41. Milentijevic, D., I.F. Rubel, A.S. Liew, D.L. Helfet, and P.A. Torzilli, An in vivo
rabbit model for cartilage trauma: a preliminary study of the influence of impact stress magnitude on chondrocyte death and matrix damage. J Orthop Trauma, 2005. 19(7): p. 466-73.
89
42. Rundell, S.A., D.C. Baars, D.M. Phillips, and R.C. Haut, The limitation of acute
necrosis in retro-patellar cartilage after a severe blunt impact to the in vivo rabbit patello-femoral joint. J Orthop Res, 2005. 23(6): p. 1363-9.
43. Szczodry, M., C.H. Coyle, S.J. Kramer, P. Smolinski, and C.R. Chu, Progressive
chondrocyte death after impact injury indicates a need for chondroprotective therapy. Am J Sports Med, 2009. 37(12): p. 2318-22.
44. Yoganandan, N., F.A. Pintar, and R. Seipel, Experimental production of extra-
and intra-articular fractures of the os calcis. J Biomech, 2000. 33(6): p. 745-9. 45. Zhang, H., M.S. Vrahas, R.V. Baratta, and D.M. Rosler, Damage to rabbit
femoral articular cartilage following direct impacts of uniform stresses: an in vitro study. Clin Biomech (Bristol, Avon), 1999. 14(8): p. 543-8.
46. Tochigi, Y., M. Rudert, D. Fredericks, A. Al-Hilli, M. Arunakul, T. Brown, and
T. McKinley, A New Impaction System to Create “Critical” Cartilage Injury in Living Rabbit Knees, in OARSI World Congress on Osteoarthritis 2011September 15-18, 2011, Osteoarthritis Cartilage: San Diego, California, USA. p. S69.
47. Backus, J.D., B.D. Furman, T. Swimmer, C.L. Kent, A.L. McNulty, L.E. Defrate,
F. Guilak, and S.A. Olson, Cartilage viability and catabolism in the intact porcine knee following transarticular impact loading with and without articular fracture. J Orthop Res, 2011. 29(4): p. 501-10.
48. Callister, W.D.J., Materials Science and Engineering: An Introduction1994, New
York, New York: John Wiley and Sons, Inc. 49. Abdel-Wahab, A.A. and V.V. Silberschmidt, Experimental and numerical
analysis of Izod impact test of cortical bone tissue. The European Physical Journal Special Topics, 2012. 206: p. 10.
50. Panagiotopoulos, E., V. Kostopoulos, S. Tsantzalis, A.P. Fortis, and A. Doulalas,
Impact energy absorption by specimens from the upper end of the human femur. Injury, 2005. 36(5): p. 613-7.
51. Zeebroeck, M.V., E. Tijskens, P.V. Liedekerke, V. Deli, J.D. Baerdemaeker, and
H. Ramon, Determination of the dynamical behaviour of biological materials during impact using a pendulum device. Journal of Sound and Vibration, 2003. 266: p. 16.
52. Tochigi, Y., T.E. Baer, P. Zhang, J.A. Martin, and T. Brown, A New Insult
Technique for a Large Animal Survival Model of Human Intra-articular Fracture, in OARSI World Congress on Osteoarthritis 2009 September 10-13, 2009, Osteoarthritis Cartilage: Montreal, Quebec, Canada. p. S51.
53. Tochigi, Y., J.A. Buckwalter, J.A. Martin, S.L. Hillis, P. Zhang, T. Vaseenon,
A.D. Lehman, and T.D. Brown, Distribution and progression of chondrocyte damage in a whole-organ model of human ankle intra-articular fracture. J Bone Joint Surg Am, 2011. 93(6): p. 533-9.